Natural Enantiomers: Occurrence, Biogenesis and Biological Properties

The knowledge that natural products (NPs) are potent and selective modulators of important biomacromolecules (e.g., DNA and proteins) has inspired some of the world’s most successful pharmaceuticals and agrochemicals. Notwithstanding these successes and despite a growing number of reports on naturally occurring pairs of enantiomers, this area of NP science still remains largely unexplored, consistent with the adage “If you don’t seek, you don’t find”. Statistically, a rapidly growing number of enantiomeric NPs have been reported in the last several years. The current review provides a comprehensive overview of recent records on natural enantiomers, with the aim of advancing awareness and providing a better understanding of the chemical diversity and biogenetic context, as well as the biological properties and therapeutic (drug discovery) potential, of enantiomeric NPs.


Introduction
Natural products (NPs) are usually regarded as small molecule organic compounds which are produced in the metabolic processes of living organisms [1]. Although studies on NPs have informed many areas of science, industry and commerce, including flavorings, perfumes, cosmeceuticals and nutraceuticals, arguably, their most important contribution to society has been as pharmaceuticals and agrochemicals [2]. For example, NPs and NPinspired chemical entities still account for more than two thirds of all the drugs approved by Food and Drug Administration (FDA) in the USA in roughly the past four decades [2].
The vast majority of reported NPs are chiral molecules that exist in nature as single enantiomers [3]. However, as the adage goes, "Beware of exceptions to the rule"; indeed, there is increasing evidence that both enantiomers of selected NPs exist in nature. Surprisingly, NPs were generally believed to exist as single enantiomers until the 1970s, despite reports of several exceptions, probably owing to the standpoint of the famous French chemist/microbiologist Louis Pasteur, i.e., that life processes were asymmetrical [4]. Benefiting from scientific and technical advances in our understanding of NP biosynthesis, there is increasing acceptance and documentation of the occurrence of natural enantiomers. Finefield et al. reported this trend in a 2012 review, documenting the occurrence and biogenesis (where applicable) of the well-known NP enantiomers reported before 2012 [3].
During our research into bioactive NPs from medicinal plants and other sources, we have regularly encountered NP enantiomers and have documented differences in their bioactivities [5][6][7][8][9]. Surveying the scientific literature revealed the aforementioned report by Finefield et al. as the only systematic record of the occurrence of natural enantiomers [3], supported by a 2018 review by Cass et al. on the techniques for separation and absolute configuration (abs. config.) assignment of enantiomeric NPs [10]. This survey also revealed a dramatic increase in the number of publications on natural enantiomers, especially in the last few years. Against this background, the present review seeks to summarize advances in this fascinating field over the period of January 2012 to December 2019.

Enantiomers from Kingdom Plantae
The kingdom Plantae is an important part of nature, providing rich resources and a beautiful environment for human beings. In the field of medicine, various plants have served as the basis of traditional herbal medication to treat a variety of diseases for thousands of years. Phytochemical research on herbs has provided thousands of structural models or leads for modern drug discovery, and some NPs can even be used directly as drugs, such as taxol. NPs derived from plants have been well studied for decades, and a comprehensive system of classification has been devised. On the other hand, new NPs from kingdom Plantae are being identified all the time due to the abundance of resources. Accordingly, enantiomers produced by plants occupy the vast majority of enantiomeric NPs from natural sources.
In this section, natural enantiomers from kingdom Plantae will be classified into fourteen subcategories on the basis of their structural type, i.e., lignans, coumarins, simple phenylpropanoids, alkaloids, flavonoids, terpenoids, phloroglucinols, naphthalene and phenanthrenes, chromanes, acetophenones, diarylheptanoids, triphenylmethanes, fatty acid and miscellaneous. Where appropriate, their biogenesis and structure will also be described.

Lignans
Lignans are a common class of NPs which is widely distributed in the plant kingdom and which exhibits a broad spectrum of bioactivities including antioxidant, antitumor, anti-inflammatory, antineurodegenerative, antiviral and antimicrobial properties [11,12]. Lignans usually consist of two (sometimes three or even more) C 6 -C 3 units (also known as phenylpropanoids). Their structural diversity arises from the different degrees of oxidation, as well as various substitution and connection patterns. Consistent with IUPAC recommendations [13], lignans are normally divided into classical lignans (only direct 8,8 -connection between the two C 6 -C 3 units), neolignans (non- 8,8 and direct connection between the two C 6 -C 3 units), oxyneolignans (ether oxygen linkage between the two C 6 -C 3 units), and higher lignans (above two C 6 -C 3 units, e.g., sesquineolignans and dineolignans). However, this classification is suggested mainly as a means of clarifying the confusing lignan nomenclature in the past, and is far from sufficient to assort the vast number of natural lignans. In general, NP chemists tend to sort lignans according to their detailed structural types, such as dibenzylbutanes, arylnaphthalenes, benzofurans, etc. [14].
Based on structural features, and for the convenience of discussion, the lignan enantiomers in the period covered by this review are classified into three subcategories: acyclic lignans, cyclic lignans and sesquineolignans. Acyclic lignans refer to those without extra rings except for the existing aromatic rings in the phenylpropanoid units, whereas cyclic lignans possess additional rings. According to the reported compound numbers, acyclic lignans are further divided into 8-4 -oxyneolignans and other acyclic lignans, while cyclic lignans will be presented as furan-incorporating lignans and other cyclic lignans. In the interests of brevity, only the structure of one enantiomer of each pair is provided; this rule applies to all structural classes in the current review.
Due to the structural flexibility of 8,4 -oxyneolignans, it has often been a challenge to correctly assign their configurations at C-7 and C-8. In order to solve this problem, three empirical rules have been developed to determine the rel. configs.: the comparison of J 7,8 coupling constants [22,23], and the utilization of 13 C (∆δ (C-8−C-7) ) [24] and 1 H (∆δ (H-9a−H-9b) ) [25] NMR chemical shift differences, although each method has its limitations.
The application of the J 7,8 value, first reported by Ruveda et al. in 1984 [23], is the simplest and most commonly used method (data see Table 1), but different substituents and their substitution positions could significantly impact the magnitude of J 7, 8 , and sometimes even result in close J 7,8 values for erythro and threo configurations. Additionally, the use of different deuterated solvents for NMR measurements will also influence the J 7,8 value. Therefore, the configuration assignments based on this empirical rule are sometimes ambiguous or even improper due to misuse. Shi and coworkers have summarized three types of 8,4 -oxyneolignans that are suitable for the application of this rule, i.e., aglycones (J 7,8 ≤ 5.0 Hz for erythro and J 7,8 ≥ 7.0 Hz for threo), aglycone acetonides (J 7,8 > 7.0 Hz for erythro and J 7,8 < 2.0 Hz for threo) and glycoside acetates (J 7,8 ≤5.3 Hz for erythro and J 7,8 ≥ 6.3 Hz for threo); the NMR solvent must be CDCl 3 [22]. As reflected by the previously reported data shown in Table 1, some researchers tend to apply this method without being aware of the aforementioned limitations, which could have resulted in incorrect configuration assignments and caused confusion in later studies of other NPs. The ∆δ (C-8−C-7) value was introduced to differentiate between erythro and threo 8,4oxyneolignans by Gan et al. [24,26], whereas only a few lignans are applied as reference compounds, and their ∆δ (C-8−C-7) values also vary in different deuterated solvents, thus causing this method to lack universality. In 2019, the third method of use of ∆δ (H-9a−H-9b) value was developed by Zhang and coworkers [25]. However, as with the rule of ∆δ (C-8−C-7) value, lack of enough model compounds has limited its application. In summary, the rel. config. determination for 8,4 -oxyneolignans could be very complicated due to their structural flexibility and diversity, and special cautions are always suggested to avoid erroneous assignments.
Up to now, three methods, i.e., direct ECD analysis by utilizing the Cotton effect at 235 ± 5 nm, TDDFT-ECD method and modified Mosher's method, have been used to establish the abs. configs. of 8,4 -oxyneolignans. For the first method, it is claimed that the positive Cotton effect at around 235 ± 5 nm is related to 8S-configuration, while a negative one corresponds to 8R-configuration [22]. However, different substituents on the aryl group, C-7, C-8 and C-9 would cause evident impact on the Cotton effects and the corresponding wavelengths. Caution thus should be taken when applying this rule, as improper applications have often been encountered in the literature. The TDDFT-ECD method is to theoretically predict the ECD spectra of the two possible enantiomers and then compare the calculated curves with the experimental ones. It is by far the most commonly used approach to assign abs. configs. of natural enantiomers owing to its easy operability, without the need for chemical derivatization and for constructing theoretical mechanisms to explain the observed properties [27]. Although this method is nowadays readily accessible to nonexperts, experience is still required since unexpected wrong assignments are easily made. As shown in Table 1, the abs. configs. determined by the first two methods are often inconsistent, and those assigned via the TDDFT-ECD method are usually accepted as the final determination in these reports. The third one is modified Mosher's method that requires chemical derivatization, and its accuracy and feasibility have been proved and accepted by almost all chemists. Nonetheless, a secondary alcohol and enough amount of sample for derivatization are a must for this method, and only pure enantiomers are suitable for investigation. In addition, it is worthwhile to note that the specific optical rotation data have no straight-forward correlation with the abs. configs. of studied structures (Table 1).
Other acyclic lignans. In addition to 8,4 -oxyneolignans, many other acyclic lignan enantiomers with various connection patterns were reported in this period, as shown in Figure 2 (names see Table S2 in Supplementary Materials). Owing to the limited numbers, they have all been put together and are discussed in the current section. Compounds 26 and 27 from Acorus tatarinowii are rare cases of naturally occurring 7,7 -oxyneolignans [28]. The threo-configurations for C-7/C-8 and C-7 /C-8 of 26 were first determined by the large J 7,8 and J 7 ,8 values (both 6.5 Hz), which was further confirmed by single crystal X-ray diffraction analysis, while the abs. configs. for 26a/26b and 27a/27b were established by comparing the calculated and experimental ECD curves. The 9,9 -oxyneolignan 28 with an ester linkage was obtained as a racemic mixture from Bulbophyllum retusiusculum [29]. Compounds 29−32 from the trunk of Torreya yunnanensis are rare examples of 8,9 -neolignans and all feature a 1,3-dioxane motif by acetalization with a 2-methoxy-cinnamaldehyde [30]. Yao and coworkers reported two 8,9 -neolignans (33 and 34) and one 7,8 -neolignan (35) as racemic mixtures from Acorus tatarinowii [28]. The J 7,8 values (7.6 Hz for 33 and 6.1 Hz for 34 in CD 3 OD) were used by the authors to determine the threo and erythro rel. configs for 33 and 34, while the rel. configs for 35 was assigned by single crystal X-ray diffraction analysis. An 8,, an 8,3 -neolignan (37) and a 7,2 -neolignan (38) were isolated from Liriodendron hybrid [31], Selaginella moellendorffii [32] and Syringa pinnatifolia [33], respectively. Sasaki et al. acquired a pair of novel 8,8 -lignan enantiomers (39a/39b) with rearranged skeleton (also known as secolignan) from Brachanthemum gobicum, with the abs. configs. being determined by comparing the ECD and specific optical rotation data with those of (−)-lucidenal [34]. The plausible biosynthetic pathway of 39 was also proposed as shown in Scheme 1, with the nonstereospecific free radical coupling being the key factor to generate enantiomers. Scheme 1. Plausible biosynthetic pathway for 39.
Benzofuran-type lignan enantiomer pairs 42−44, 45 and 46 were discovered from Jatropha integerrima in 2015 [39], Brachanthemum gobicum in 2018 [34] and Picrasma quassioides in 2018 [40], respectively. The 7,8-trans configurations for 42−46 were determined by the large J 7,8 values (7.5 Hz for 42−45 in CDCl 3 , and 6.7 Hz for 46 in DMSO-d 6 ) and NOE analyses, while the abs. config. assignments for these compounds were based on the reversed helicity rule [39,41]. According to this empirical rule, P-helicity of the nonaromatic ring will lead to a positive Cotton effect within the 1 L b band (around 280 nm) and M-helicity will result in a negative Cotton effect. Phytochemical investigation into the plants Rubus idaeus [42,43] and Phyllanthus glaucus [44] led to the isolation of three enantiomer pairs 47−49, and their abs. configs. were established by TDDFT-ECD method. Compounds 50−52 incorporating an α,β-unsaturated aldehyde unit were obtained from Brachanthemum gobicum [34] and Picrasma quassioides [40], with the abs. configs. also being assigned by the reversed helicity rule, where their 1 L b (α) bands red shifted to around 340 nm due to the conjugation effect from the α,β-unsaturated aldehyde group. In 2018, Huang et al. discovered compound 53 as a racemic mixture from Rubus ideaus and further resolved it into two enantiomers (53a/53b), with the abs. configs. being assigned by application of the TDDFT-ECD method [43]. In fact, compound 53 had been previously reported as an optically pure molecule from Broussonetia papyrifera in 2009, with a much smaller [α] D value [45], suggesting its potential scalemic nature. Compounds 54−57 are a group of dinorneolignans and were isolated from Rubus idaeus [42,43] and Brachanthemum gobicum [34].
In 2019, Song and colleagues obtained the trinorneolignan furolactone 58 as a racemic mixture from Rubus idaeus and resolved it into a pair of enantiomers (58a/58b), the abs. configs. of which were established by analyses of the calculated shielding tensor values and ECD data [46], while the enantiomer 58b had been reported as an optically pure molecule from Lycium chinense in 2013 [47]. The other two pairs of furolactone enantiomers 59a/59b and 60a/60b were isolated from Archidendron clypearia in 2018 [48] and Dendrobium nobile in 2016 [49], respectively. Song and coworkers reported 61a/61b, from Rubus idaeus in 2019 and assigned their abs. configs. by using the TDDFT-ECD method [47]. Compounds 62 and 63 represent another two pairs of furofuran-type lignan enantiomers isolated from Morinda citrifolia [50] and Acorus tatarinowii [16], respectively.
Other cyclic lignans. Except for the aforementioned Furan-incorporating lignan enantiomers, there are also some other cyclic lignan enantiomers with diverse ring systems as listed in Figure 4 (names see Table S4 in Supplementary Materials). The unusual dinorneolignans 64a/64b incorporating a 1,4-dioxane motif and the arylnaphthalene-type lignans 67a/67b bearing a 2 ,9 -epoxy ring were separated from Pithecellobium clypearia in 2018, with their abs. configs. being determined by the TDDFT-ECD method [51]. In 2015, Zhang et al. reported a pair of 7,8 -epoxy-8,7 -oxyneolignans (65a/65b) and a pair of oxidized arylnaphthalene-type lignans (68a/68b) from Acorus tatarinowii, and the abs. configs. of the two pairs were established by employing the TDDFT-ECD method and comparing the Cotton effect at 315 nm with those of known analogues, respectively [38]. The other pair of enantiomeric arylnaphthalenes 66a/66b were also isolated from Acorus tatarinowii, with their abs. configs. being assigned by comparing the Cotton effect at 285 nm with those of known analogues [28]. Compounds 69a/69b featuring a cyclobutane ring via 7,7 and 8,8 connections represent a pair of [2 + 2] cycloaddition adducts of two phenylpropanoid units and were obtained from Isatis indigotica in 2019 [52]. Isolated from Tylopilus eximius in 2012 are two pairs of enantiomers 70a/70b and 71a/71b both incorporating a cyclopentenone ring formed by 7,8 and 9,7 linkages [53]. The rel. config. of racemic 71 was established by X-ray crystallography, while the abs. configs. of 70a/70b and 71a/71b were confirmed by TDDFT-ECD method. In 2015, two pairs of rare spirodienone neolignans (72a/72b and 73a/73b) were reported from Cinnamomum subavenium, with the absolute structures being elucidated by X-ray crystallographic analysis and TDDFT-ECD method [54].

Sesquineolignans
Sesquineolignans refer to lignans bearing three phenylpropanoid units with various connection patterns. Ten pairs of enantiomeric sesquineolignans were reported in this period, and their structures are shown in Figure 5, with the names being listed in Table S5 in Supplementary Materials. In 2015, Yin and coworkers reported 79 and 80 from Brachanthemum gobicum and applied the reversed helicity rule to assign the abs. configs. at C-7 and C-8 [34], while the assignments of abs. configs. at C-7" and C-8" were established by Rh 2 (OCOCF 3 ) 4 -induced ECD analysis. On the basis of the bulkiness rule for secondary alcohols, a positive Cotton effect at around 350 nm (E band) in the Rh 2 (OCOCF 3 ) 4 -induced ECD spectrum indicated a S-configuration, while a negative Cotton effect implied a R-configuration. In 2014, Yu's group discovered a pair of novel enantiomeric tetrahydrofuran spirodienone sesquineolignans (81a/81b) from Xanthium sibiricum and proposed coniferyl alcohol as the biosynthetic precursor (Scheme 2), and the nonstereospecific radical coupling between the two C 6 -C 3 units was the key factor to result in enantiomers [55]. Song and coworkers reported 82 and 83 from Rubus idaeus in 2019 and assigned their abs. configs. by using the TDDFT-ECD method [47].
In summary, a large number of lignans (except lignan glycosides) have been discovered as racemic or scalemic mixtures and chirally separated in recent years, and their structural types, from simple to complex (via rearrangement), cover more than half of the known classes. It is self-evident from the aforementioned examples that enantiomerism widely occurs in the structural categories of lignans especially for 8,4 -oxyneolignans and furan-incorporating lignans. These lignan enantiomers exist as either racemic or scalemic mixtures in plants and can be relatively easily separated by commercially available chiral chromatographic columns. Therefore, it is conceivable that many examples previously reported as optically pure lignans could in fact be scalemic mixtures, and NP researchers should pay extra attention to the enantiomeric purity of lignans in their future work. Scheme 2. Plausible biosynthetic pathways for 81a/81b.

Coumarins
'Coumarin' is the general name of ortho-hydroxycinnamate lactones that are derived from the Shikimate biosynthetic pathway 1 (for structures, see Figure 6; for names, see Table S6 in Supplementary Materials). As the core backbone of coumarins does not contain chiral factors, the generation of their enantiomersim usually comes from the chiral carbons of substituents (e.g., prenyl substitution) or axial chirality of oligomers. Coumarins are important secondary metabolites in plants and have shown various biological properties such as antitumor, anti-HIV, antimicrobial and anti-inflammatory activities [56,57].
Compounds 84a−87a are a group of angular dihydropyranocoumarins and were obtained as 3 S,4 S-configured pure enantiomers from Peucedanum japonicum in 2017 [58], while their corresponding 3 R,4 R-enantiomers (84b−87b) had been previously reported from Angelica morii in 1974 [59], Peucedanum praeruptorum in 2012 [60], Seseli gummiferum in 1971 [61] and Angelica furcijuga in 2000 [62], respectively. Another eight pairs of analogues 88−95 were found to be present as scalemic mixtures in Peucedani Radix [63] and were successfully separated into pure enantiomers for the first time. Except 88a, 89b, 92a and 92b, the others have been formerly reported as optically pure compounds [63], but the small [α] D values compared with those for the purified enantiomers suggested their scalemic natures. Tang and coworkers isolated two pairs of coumarin enantiomers (96a/96b and 97a/97b) from Toddalia asiatica and assigned the rel. and abs. configs. of 96a/96b via X-ray diffraction experiment and TDDFT-ECD method, respectively [64]. From Sapium baccatum, three coumarin enantiomer pairs incorporating one additional α-pyrone ring (98−100) were assigned the abs. configs. by comparing their specific optical rotations with those of known analogues [65]. Compounds 102 and 103 are two pairs of hybrid dimer enantiomers from Cnidium monnieri and they were also total synthesized for further biological test [66]. The most complex coumarin enantiomers so far are the oligomeric coumarin hybrids 104 and 105 bearing a spirodienone-sesquiterpene skeleton, and they were isolated from Toddalia asiatica in 2016 [67]. The only coumarin enantiomers generated by axial chirality are the prenylated coumarin dimers 101a/101b, with the abs. configs. being determined by TDDFT-ECD method [68].

Simple Phenylpropanoids
Simple phenylpropanoids are naturally occurring phenolic substances containing only one C 6 -C 3 biosynthetic block. They often exist as racemic or scalemic mixtures in nature (for structures, see Figure 7; for names, see Table S7 in Supplementary Materials). They are also derived from the Shikimate biosynthetic pathway [1]. Many phenylpropanoids play vital roles in plant growth regulation and pathogen defense by acting as essential components of cell wall, as protectants against high light and UV radiation, and as phytoalexins against herbivores and pathogens [69]. It is generally difficult to acquire high quality crystals for X-ray diffraction analysis due to the rotary nature of the sidechains in most phenylpropanoids, so normally their rel. configs. are assigned by J values and the abs. configs. are determined on the basis of Snatzke's rule, modified Mosher's method or TDDFT-ECD calculation.
Qiu and coworkers reported two pairs of phenylpropanoid enantiomers, 106a/106b and 107a/107b, from the leaves of Eucommia ulmoides and assigned their rel. and abs. configs. by analysis of J 7,8 values and Snatzke's method, respectively [70]. The planar structure of 108 had already been reported in 2001 [71], but its enantiomeric nature was not revealed by Liu et al. until 2017, with the abs. configs. being determined by Snatzke's rule [72]. Two pairs of rare chlorine-containing enantiomers (109a/109b and 110a/110b) were isolated from Acorus tatarinowii in 2017, and their rel. and abs. configs. were established by analyzing J 7,8 values and employing modified Mosher's method, respectively [73]. The enantiomer pairs 111a/111b−115a/115b were obtained from Acorus tatarinowi in 2017 by Gao's group, among which 111b and 113b had been reported previously [74]. Compounds 115a/115b are the first cases in nature of asarone-derived phenylpropanoids with an isopropyl fragment tethered to the benzene core, and their abs. configs. were assigned by TDDFT-ECD method [74]. Song and colleagues isolated 116a/116b and 117a/117b from the fruit of Crataegus pinnatifida in 2018 and applied the TDDFT-ECD method to establish their abs. configs [74]. Compounds 118−126 with an extra phenyl group on the sidechain are also considered to be 1,2-diphenylpropane derivatives. The enantiomer pairs 118−120 were separated from Rubus idaeus in 2018 with their abs. configs. being determined via the TDDFT-ECD method [75]. Compounds 121 and 122 were first reported as racemic mixtures from Casearia grewiifolia in 2012 [76] and were resolved into two pairs of enantiomers by Qiu et al. in 2018, with the rel. and abs. configs. being determined by analyzing J 7,8 values and applying TDDFT-ECD method, respectively [77]. Compounds 123a/123b−126a/126b featuring a 1,3-dioxane ring derived from condensation of diol with different aldehydes, were obtained from Crataegus pinnatifida with their abs. configs. being determined by TDDFT-ECD method [78].

Alkaloids
The term "alkaloids" traditionally describes nitrogen-containing small molecule organic compounds with basicity, although there is no unified definition. In this section, we include all nitrogen-bearing NPs in this category. Based on the structural types, natural alkaloid enantiomers from plants in the period of 2012-2019 are classified into indole alkaloids, quinoline and isoquinoline alkaloids, β-carboline and carbazole alkaloids, piperidine alkaloids, thiohydantoin alkaloids, indolizidine and quinolizidine alkaloids, and other alkaloids.
It is interesting to note that suitable crystals for X-ray diffraction analysis of the enantiomeric mixtures seem relatively easy to be obtained in these reports, but the acquisition of high quality crystals of pure single enantiomers appears difficult. As above described, from simple indoles (e.g., 127) to monoterpenoid indole hybrids (e.g., 140), from single indoles (e.g., 129) to dimeric indoles (e.g., 134), from one-chiral-center examples (e.g., 131) to complex multiple-chiral-center indole dimers (e.g., 144), natural indole alkaloid enantiomers spread in a wide range of structural subtypes. Therefore, checking enantiomeric purity for this important class of NPs appears to be key in the future work.

Quinoline and Isoquinoline Alkaloids
Most quinoline and isoquinoline alkaloids biosynthetically originate from anthranilic acid or from indoles via rearrangement [94]. Quinoline & isoquinoline alkaloids have attracted great interest from researchers worldwide because of their wide-range biological activities, including antitumor, antiparasitic and insecticidal, antibacterial and antifungal, cardioprotective, antiviral, antiinflammatory, hepatoprotective, antioxidant, anti-asthma, antitussive, and other activities [95,96]. The structures and names of these alkaloid enantiomers in the covered stage are summarized in Figure 9 and Table S9 in Supplementary Materials, respectively.
Compounds 147a/147b, featuring a furoquinoline core hybridized with a phenylpropanoid unit via a 1,4-dioxane ring, were separated and characterized from Zanthoxylum nitidum in 2018 [97]. Three 2-quinolinone enantiomer pairs 148−150 were reported from Isatis tinctoria by Song and coworkers in 2019 [81], and in the same year, Zhang et al. discovered the same type of alkaloid enantiomers 151a/151b from the roots of Isatis indigotica [86]. The last example of quinolinone enantiomer pair is compound 152 also from I. indigotica, and it incorporates an additional anthranilic acid residue [98].
Compounds 153−158 are a series of isoquinoline enantiomers, among which 154a/154b were acquired from Corydalis hendersonii in 2016 [99] and the others were obtained from Corydalis mucronifera in 2018 [100]. Compounds 154a/154b were proposed to be derived from the condensation of a benzylisoquinoline and a succinic acid [99]. In 2016, Hua and colleagues reported from Macleaya cordata five dihydrobenzophenanthridine enantiomer pairs 159−163 and a racemate 164, among which 162 and 163 had been previously isolated in racemic form from Macleaya cordata [101] and here is the first record of their chiral separation [102]. Three same type of enantiomer pairs 165, 166 and 173 were isolated and characterized from Corydalis ambigua var. amurensis by Han and coworkers, and three racemic mixtures 167−169 were also acquired and analyzed by chiral chromatography but without further separation due to their limited amount [103]. As for structure elucidation, single-crystal X-ray diffraction analysis was applied to determine the abs. config. of 165a, followed by the abs. config. assignments for 165b and 166a/166b via comparing the ECD curves with that of 165a. In addition, Ye's group reported a pair of berberine-type alkaloid enantiomers 170a/170b from Coptis chinensis in 2014 [104]. Sai et al. discovered from Corydalis ambigua two pairs of alkaloid dimers 171a/171b and 172a/172b, featuring a novel dimerization pattern from two different types of monomers via a C-C single bond [105]. The plausible biosynthetic pathways for 171 and 172 were also proposed as shown in Scheme 3 by the authors, and the nonstereospecific nucleophilic addition was assumed to be the key factor to generate enantiomers [105]. As with the aforementioned indole alkaloids, the assignments of abs. configs. for most quinoline and isoquinoline enantiomers have been based on the TDDFT-ECD method.

β-Carboline and Carbazole Alkaloids
β-Carbolines and carbazoles are among the most intriguing alkaloid groups; they derive from various sources. They have gained increasing attention due to their broad spectrum of biological activities [106,107]. Seven β-carboline (174, 178−183), three βcarboline-carbazole hybrid (175−177) and nine carbazole (184−192) enantiomer pairs have been reported in this period (for structures, see Figure 10; for names, see Table S10 in Supplementary Materials). The abs. configs. for all separated enantiomers in this section were determined by the TDDFT-ECD method unless otherwise specified.

β-Carboline and Carbazole Alkaloids
β-Carbolines and carbazoles are among the most intriguing alkaloid groups; they derive from various sources. They have gained increasing attention due to their broad spectrum of biological activities [106,107]. Seven β-carboline (174, 178−183), three βcarboline-carbazole hybrid (175−177) and nine carbazole (184−192) enantiomer pairs have been reported in this period (for structures, see Figure 10; for names, see Table S10 in Supplementary Materials). The abs. configs. for all separated enantiomers in this section were determined by the TDDFT-ECD method unless otherwise specified. Song and coworkers phytochemically studied the stems of Picrasma quassioides to detect four enantiomer pairs 174a/174b−177a/177b. While 174a/174b possess a β-carbolinephenylpropanoid hybrid skeleton [108], the latter three pairs represent alkaloid heterodimers of a β-carboline and a carbazole units which are linked via a C 4 fragment. Alkaloids 178a/178b−180a/180b are dimeric β-carbolines obtained as trifluoroacetates from Picrasma quassioides in different years [109,110]. Compounds 181a/181b, as β-carbolinequinazoline hybrid dimers from Peganum harmala, were biogenetically produced through Mannich/Pictet-Spengler-type and intermolecular Michael addition reactions [111]. Compounds 182 and 183 from Pausinystalia yohimbe were characterized in racemic forms in 2018 without further chiral separation, and their racemic nature was further proved by X-ray diffraction analysis [112]. Interestingly, the enantiomerism of 182 results from the N-4 chiral center which is very rare in nature [112].
The enantiomerism of carbazole alkaloids comes from the axial chirality of dimers or from the chiral centers in the additional structural fragments. Four pairs of biscarbazole atropisomers (184a/184b−187a/187b) and a pair of dihydropyranocarbazole enantiomers (188a/188b) were discovered by Jiang and colleagues from Clausena dunniana, where the planar structure of 185 had been previously described from Clausena wallichii in 2011 [113]. The same authors from Jiang's group further reported 189a/189b−192a/192b from Murraya microphylla [114,115], with the rel. config. of 189 being confirmed by X-ray crystallographic data [115].

Piperidine Alkaloids
Piperidine alkaloids that have one or more piperidine rings in the structures are generally believed to be biogenetically derived from lysine [1]. During the period covered by this review, fifteen pairs of piperidine enantiomers (for structures, see Figure 11; for names, see Table S11 in Supplementary Materials) have been reported. The enantiomer pairs 193−197 were isolated from Anacyclus pyrethrum in 2018 [116]. Among them, compounds 193 and 194 possess novel dimeric piperidine backbones with 6/5/6/6 and 6/5/6 ring systems, respectively, while 195a/195b incorporate a rare cyclopentane-piperidine framework. In 2017, compounds 198−205 were obtained from Viola tianschanica. All of them bear more than one nitrogen atom and incorporate fascinating heterocyclic architectures such as the 6/5/6/5 and 6/5/5/6/5 ring systems in 198 and 201-202, respectively [117]. The abs. configs. of these alkaloid enantiomers were established by the TDDFT-ECD method. Compounds 206 and 207 were found to occur as enantiomeric pairs in Clausena lansium with only 206 being successfully resolved into pure enantiomers. The rel. and abs. configs. of 206a/206b were established by X-ray crystal data and comparing ECD and specific optical rotation data with calculated ones [118].
The biogenetic origins of these piperidines, especially of those with highly complex skeletons and multiple chiral centers like 198 and 201-202, have not been examined, and this intriguing puzzle definitely deserves further investigations.

Thiohydantoin Alkaloids
Naturally occurring thiohydantoin alkaloids are a rare class of NPs. Compounds 208a/208b−218a/218b (for structures, see Figure 12, names see Table S12 in Supplementary Materials), a panel of thiohydantoin derivatives of two structural groups, were initially obtained as racemic mixtures from Lepidium meyenii and further resolved into eleven pairs of enantiomers [119]. Among them, an unidentified enantiomer of 208 had been reported as a synthetic product in 2007 [120]. Although the biogenesis of these alkaloids has never been studied, they very likely belong to the imidazole class originating from histidine on the basis of their core structures [1].

Indolizidine and Quinolizidine Alkaloids
Both indolizidine and quinolizidine alkaloids are biogenetically originated from lysine [1], and an equal number of four pairs of indolizidine (219−222) and quinolizidine (223−226) enantiomers (for structures, see Figure 13; for names, see Table S13 in Supplementary Materials) have been reported in the period covered by this review. Alkaloid 219 from Ficus fistulosa var. tengerensis was identified as a scalemic mixture by [α] D , ECD and X-ray crystallographic data [121], while 220 as a long-known NP reisolated from Tylophora indica [122] was demonstrated to be a nearly racemic mixture with only a slight excess of the R-enantiomer [123]. Compounds 221a/221b, a pair of enantiomeric indolizidine alkaloid dimers from Dendrobium crepidatum, were assigned the abs. configs. by single-crystal X-ray diffraction analysis [124]. Enantiomers 222a/222b, whose structures were also confirmed by X-ray diffraction analysis to be indolizidine dimers linked via a cyclobutane ring, were obtained from the same species as 219 [121]. Zhang et al. discovered four pairs of neosecurinane-type alkaloid enantiomers 223a/223b−226a/226b of the quinolizidine class from Flueggea virosa in 2017, and it is the first time to report the enantiomerism of this interesting type of alkaloids [5]. The rel. and abs. configs. of 223a/223b−226a/226b were characterized by a variety of techniques including X-ray crystallography and ECD experiments.

Other Alkaloids
In addition to the aforementioned alkaloid enantiomers occurring naturally in plants, there are also many other types of alkaloid enantiomers reported in this period, as summarized in Figure 14 (names see Table S14 in Supplementary Materials). Compounds 227a/227b−230a/230b are a panel of quinazoline enantiomer pairs obtained from Peganum harmala in 2018 [125], Isatis indigotica in 2019 [86], I. indigotica in 2016 [98] and P. harmala in 2016 [126], respectively. Biogenetically, quinazoline alkaloids have been demonstrated to be derived from anthranilic acid [1].
Except for the specified ones, the abs. configs. of the alkaloid enantiomers in this section were all established by applying the TDDFT-ECD method.

Flavonoids
Flavonoids are a large family of secondary metabolites that exist widely in the plant kingdom. They exhibit a variety of bioactivities such as anti-inflammatory, antioxidant, antibacterial, antiviral, anticancer and neuroprotective effects [135]. Traditionally, flavonoids mainly refer to compounds incorporating a 2-phenylchromone core, and nowadays, this term has extended to all structures with two phenyl units linked via a C 3 fragment [14]. In addition, some NPs such as xanthones and furanochromones are also included in this structural family as atypical flavonoids. Flavonoid enantiomers reported in this period are classified into three subgroups: flavones and isoflavones, chalcones and xanthones

Flavones and Isoflavones
In this section, the definition 'flavones' refers to all those incorporating the basic 2phenylchromone backbone, including classical flavones, flavanones, flavanes, etc. The abs. configs. of these enantiomers are mostly determined by the TDDFT-ECD method unless otherwise specified. Their structures and names are shown in Figure 15 and Table S15 in Supplementary Materials, respectively.
As can been from the above-mentioned structures, the enantiomerism of these flavones mainly comes from either the chirality of flavanone/flavane core or that of additional structural units especially prenyl group(s), or both. Meanwhile, the enantiomerism of the described isoflavones arises exclusively from the chirality of extra structural units, i.e., prenyl group(s) and phenylpropanoid fragment for the current cases.

Xanthones
Xanthones are polyphenolic compounds incorporating a common 9H-xanthen-9-one scaffold with various substituents, making them 'privileged structures' which are likely to bind to a variety of biological targets. They have been shown to display significant bioactivities including antimicrobial, antioxidant, cytotoxic activities, and so on [155]. Most xanthone enantiomers reported in this period are prenylated; their structures are listed in Figure 17 (names see Table S17 in Supplementary Materials). Hua and coworkers reported a pair of diprenylated xanthone enantiomers 297a/297b with only one chiral center from Cratoxylum cochinchinense in 2019 [156]. The deoxoxanthone enantiomers 298a/298b and 299a/299b incorporating a phenylpropanoid unit were isolated from Uvaria valderramensis in 2014 [157]. Also in 2014, three pairs of prenylxanthone enantiomers (300a/300b−302a/302b) were isolated from Cratoxylum formosum, with the abs. configs. being established by X-ray crystallographic experiment [158]. As shown in Scheme 4, the generation of enantiomeric 300a/300b−302a/302b is plausibly derived from diallylxanthone through a key process of nonstereospecific Claisen rearrangement [158]. In addition to 300−302, eleven pairs of caged prenylxanthone enantiomers 303a/303b−307a/307b and 308a/308b−313a/313b were reported from Garcinia bracteata in 2018 [159] and from Garcinia propinqua in 2017 [160], respectively, with the abs. config. of 313a being determined by single-crystal X-ray diffraction analysis. The biogenetic origins of those xanthones with multiple chiral centers are indeed interesting topics that deserves further investigations. Scheme 4. Plausible biosynthetic pathways for 300a/300b−302a/302b.

Terpenoids
Terpenoids are probably the biggest family of NPs with diverse structures and various biological activities [3]. All terpenoids are initially assembled from the head-to-tail condensation of repeated isoprene units (C 5 ), and according to the number of isoprene residues, terpenoids are normally classified into monoterpenoids (C 10 ), sesquiterpenoids (C 15 ), diterpenoids (C 20 ), sesterterpenoids (C 25 ) and triterpenoids (C 30 ). Additionally, meroterpenoids are also an interesting class of terpenoid products with mixed biogenesis [3]. To the best of our knowledge, enantiomeric cases have been reported for all terpenoid subclasses except triterpenoids.

Sesquiterpenoids
Sesquiterpenoids are constructed from three isoprenyl fragments. Among all the terpenoid classes, they have the most diverse carbon skeletons and are probably the largest group of terpenoid NPs. Corresponding to their various structural types, natural sesquiterpenoids have also exhibited a myriad of biological properties [161], and this has been well reflected by the success of artemisinin (for malaria), the most famous sesquiterpenoid whose discovery was rewarded the Nobel prize in Physiology or Medicine in 2015. Enantiomeric sesquiterpenoids reported in this period include 16 pairs (314−329) with different backbones (for structures, see Figure 18, names see Table S18 in Supplementary Materials).

Diterpenoids
Diterpenoids are biosynthesized from the head-to-tail condensation of four isoprene (C 20 ) units, and have the second largest number of carbon backbones in the terpenoid family. Like sesquiterpenoids, they are also well-known in the NP community for their diverse bioactivities particularly antitumor effects with therapeutic values [168]. As is well known, the most notable diterpenoid is taxol, which has been used as a cancer treatment for over three decades. Nine pairs of enantiomeric diterpenoids 330a/330b−338a/338b (for structures, see Figure 19, names see Table S19 in Supplementary Materials) with different structural skeletons were recorded in the study period. Yue and coworkers phytochemically investigated Croton mangelong to afford two pairs of macrocyclic diterpenoid enantiomers 330a/330b and 331a/331b featuring a bicyclo[9.3.1]pentadecane core and a rare bridgehead double bond, with the abs. config. for 330b being determined by single-crystal X-ray diffraction analysis [169]. Compounds 332a/332b are bis-seco-abietane diterpenoids from Cryptomeria fortune and were asymmetrically synthesized through a readily made intermediate orthoquinone from sugiol [170]. Compounds 333a/333b are a pair of norditerpenoid enantiomers from the roots of Salvia miltiorrhiza [171]. Compounds 334a/334b, rearranged abietane-type diterpeniods featuring a 5/6/6 tricyclic architecture with the five-membered ring formed via C-2-C-11 single bond, were isolated and characterized from Salvia prionitis in 2015 [171]. Compounds 335a/335b are a pair of diterpenoid enantiomers with a highly oxygenated novel backbone obtained from Swertia leducii in 2014, with the rel. config. being determined by X-ray diffraction analysis and the abs. configs. by TDDFT-ECD method [172]. In the same year, compounds 336a/336b were isolated from Paeonia veitchii [173], and they incorporate an aromatized norditerpenoid skeleton, with the rel. config. being confirmed by X-ray crystallography. Compounds 337a/337b and 338a/338b, two pairs of norditerpenoid enantiomers with unusual 5,5-spiroketal core, were obtained from Hypericum japonicum in 2016, with the abs. configs. being assigned by a combination of TDDFT-ECD calculation, modified Mosher's method and quantum chemical predictions (QCP) of 13 C NMR data [174].

Phloroglucinols
Phloroglucinol derivatives represent a unique class of NPs featuring one or more intact or modified phloroglucinol units, with alkylation and acylation as the common structural modifications [181]. The chirality of them arises usually from their prenyl/terpenyl substituents and/or from the dearomatization of phloroglucinol core. In most cases, they can also be classified into the 'meroterpenoid' group, but here we describe them separately owing to the considerable number of reports and their popularity among NP workers in recent years. Phloroglucinal enantiomers reported in this period all incorporate one or more acyl groups including acetyl, isobutyryl, benzoyl, cinnamoyl and dihydrocinnamoyl, and their structures are shown in Figure 21 (names see Table S21 in Supplementary Materials). Wherever the abs. configs. are determined by the TDDFT-ECD method, it will be not specifically mentioned in this section.

Naphthalenes and Phenanthrenes
The enantiomerism of naphthalene and phenanthrene derivatives is generally attributable to chiral centers and, in many cases, axial chirality. The structures of the naphthalene and phenanthrene enantiomers reported in this period are displayed in Figure 22 (names see Table S22 in Supplementary Materials). The abs. configs. of most of these enantiomers were determined by the TDDFT-ECD method, unless otherwise specified.

Naphthalenes and Phenanthrenes
The enantiomerism of naphthalene and phenanthrene derivatives is generally attributable to chiral centers and, in many cases, axial chirality. The structures of the naphthalene and phenanthrene enantiomers reported in this period are displayed in Figure 22 (names see Table S22 in Supplementary Materials). The abs. configs. of most of these enantiomers were determined by the TDDFT-ECD method, unless otherwise specified. configs. of 386 and 388 being confirmed by X-ray diffraction analysis and 13 C NMR calculation, respectively [8]. Compounds 392−394, three pairs of 3,4-dihydro-4-naphthylnaphthalen-1(2H)-one enantiomers, were obtained from Juglans regia in 2019 [193]. Compounds 395a/395b from Rubia oncotricha were characterized by Tan and coworkers as novel naphthoquinone dimers with an unprecedented spiro [4.5] carbon core [194]. Another pair of dimeric naphthoquinone enantiomers 396a/396b were also reported by the same research team from Rubia alata in 2014, with the rel. config. being corroborated

Phenanthrenes
Phenanthrene enantiomers (398a/398b−408a/408b) in this period have been solely reported from Bletilla striata by Li's and Hou's research teams in 2019 [130], and they can be divided into three groups, namely, phenanthrene monomer (405), phenanthrene dimers (404 and 406−408) and phenanthrene-phenylpropanoid hybrids (398−403). The enantiomerism of these compounds has been generated from the axial chirality of phenanthrene moiety and/or from the chiral centers of phenylpropanoid unit. Notably, compounds 404−408 with only axial chirality were able to be separated into five pairs of enantiomers. Axial chirality, although well known to organic chemists, has often been overlooked by NP researchers, owing to its rare presence in natural molecules. Therefore, the enantiomeric purity of NPs with axial chirality is strongly recommended to be checked no matter they are new or known.

Chromanes
Chromane derivatives are a class of NPs having a chromane core or a modified one (e.g., chromone, chromanone) in their structures. Enantiomeric chromane derivatives reported in this period are listed in Figure 23 (names see Table S23 in Supplementary Materials). Compounds 409a/409b−415a/415b had been studied previously in many occasions as pure enantiomers, scalemic mixtures or racemates, and as natural molecules, biotransformation products or synthetic intermediates, but none of these reports had paid attention to the enantiomerism of this group of structures. They were separated from the flower buds of Tussilago farfara in the authors' lab in 2018, with the abs. configs. being determined by chemical method as well as TDDFT-ECD calculation and ECD comparison [7]. Compounds 416a/416b were proposed to be a pair of norbisabolane sesquiterpenoid enantiomers yet incorporating a chromone core and were obtained from Curcuma longa in 2019 [197]. Compounds 417a/417b−421a/421b are five pairs of prenylated chromone enantiomers isolated from Harrisonia perforate in 2014, whereas only the abs. configs. of 417a/417b were assigned by Mosher's method [198]. From the same plant, 422a/422b were reported as a pair of enantiomeric molecules by Yuan et al. in 2017 [199].

Diarylheptanoids
Diarylheptanoids, a class of NPs characteristic of a 1,7-diphenylheptane core, have been increasingly recognized as potential therapeutic agents for their diverse biological properties including antiinflammatory, antitumor, antioxidant, antiestrogen, hepatoprotective, antileishmania and neuroprotective activities [205]. Nine pairs of diarylheptanoid enantiomers (433−441, for structures, see Figure 25; for names, see Table S25 in Supplementary Materials) have been documented in the covered period. The occurrence of enantiomerism in these compounds results from the chiral centers generated by oxidation or Diels-Alder cycloaddition with other molecules. The enantiomeric pairs of diarylheptanoid-monnoterpene adduct (433 & 434) and diarylheptanoid-sesquiterpene hybrid (435-437) from Alpinia officinarum, were hypothesized to be produced via a crucial Diels-Alder cycloaddition between the diarylheptanoids and corresponding terpenyl units. The rel. configs. for the chiral centers in the cyclohexene ring were assigned by comparing the experimental and calculated 13 C NMR data, followed by the establishment of the abs. configs. via the TDDFT-ECD method [206]. Compounds 438−441 are four pairs of diarylheptanoid enantiomers acquired from Dioscorea villosa in 2012, with the abs. configs. being determined by the modified Mosher's method [207].

Triphenylmethanes
Triphenylmethanes are a unique class of NPs with one central carbon being linked by three aryl groups. They have been discovered to have a wide range of biological activities including antioxidant, antitumor, anti-HPK (histidine protein kinases) activities, etc. [208]. Six pairs of triphenylmethane enantiomers (442−447, Figure 26, Table S26 in Supplementary Materials) were reported in this period; their enantiomerism is attributable to chiral centers (442 and 443) or axial chirality (444−447). Compounds 442a/442b and 443a/443b are two pairs of triarylmethane enantiomers reported from Securidaca inappendiculata in 2018, with the abs. configs. being determined by X-ray crystallography [209]. In addition, bio-inspired total syntheses for these compounds were also completed [209]. Compounds 444−447 occurred as racemates generated by axial chirality in the plant Selaginella pulvinate [210], and subsequent chiral fractionation divided 444 and 447 into 444a/444b and 447a/447b, respectively, with the abs. configs. being assigned by TDDFT-ECD method. However, 445 and 446 had not been enantiomerically separated.

Fatty Acids
Five pairs of enantiomeric fatty acid esters (448a/448b−452a/452b) were recorded in this covered stage and their structures are listed in Figure 27, with names being shown in Table S27 in Supplementary Materials. Usually, the generation of chirality in these compounds derives from the nonstereoselective oxidations on the aliphatic chain (448−451) or substitution on the glycerol moiety (452). Compounds 448a/448b−452a/452b from Plantago depressa were characterized as four pairs of 9-oxo octadecanoid derivatives by the authors' group, with 451 bearing a rare chlorine atom [6]. We have also isolated 452a/452b as octadecanoid monoglycerides from the seeds of Ipomoea nil in 2019 and established their abs. configs. via an in situ dimolybdenum ECD method [9].

Miscellaneous
Other enantiomeric NPs from plants reported in this period are displayed in Figure 28 (names see Table S28). The abs. configs. of all these enantiomers were assigned by the TDDFT-ECD method unless otherwise specified. Compounds 453−456 are four pairs of enantiomeric phthalide derivatives, all of which were isolated and characterized from Angelica sinensis in 2018 [211]. Phthalides are a rare class of NPs referring to lactones of 2-hydroxymethyl benzoic acids. They exist in nature as monomers or oligomers, and the latter are generally produced via [2 + 2] or [4 + 2] cycloaddition to form a number of complex polycyclic skeletons with multiple chiral centers [211]. Among them, 453 and 454 are dimers, while 455 and 456 are trimers.
Compounds 457a/457b and 458a/458b are enantiomeric stilbenoids that have 1,2diphenylethylene (stilbene) as their basic scaffold and exist as monomers or oligomers in nature. They normally act as phytoalexins to assist plants in their resistance to pathogens or stress factors [212]. Compounds 457a/457b, prenylated stilbenoid dimers isolated from Cajanus cajan in 2014, possess an interesting dimerization pattern generated from nonstereoselective radical addition as shown in Scheme 7, and their structures including the abs. configs. were determined by a combination of X-ray diffraction analysis and TDDFT-ECD calculation [213]. Compounds 458a/458b are enantiomeric stilbenoid trimers obtained from Cyperus rhizomes in 2012, and their abs. configs. were established by comparing the [α] D and ECD data with those of known analogues [214,215].

Enantiomers from Kingdom Fungi
Enantiomers originating from fungi, i.e., from phyla Ascomycota and Basidiomycota, will be presented in this section. The structural classification of NPs from fungi is not as regular and clear as those from plants; a myriad of fungal NPs belong to the super family of polyketides that derive biogenetically from the acetate pathway [1]. Also, considering the limited number of molecules described in this section, the enantiomers described here are simply divided into nonalkaloids and alkaloids. Where applicable, their biogenesis and structure will also be described.

Alkaloids
Alkaloid enantiomers found in phylum Ascomycota during the reported period are displayed in Figure 30, with their names being shown in Table S30 in Supplementary Materials. The abs. configs. of those established by TDDFT-ECD method are not specifically mentioned in this section.

Enantiomers from Phylum Basidiomycota
It is interesting to note that all natural enantiomers from phylum Basidiomycota collected in this period, with only one exception (Granulobasidium vellereum), were reported from species of the well-known medicinal macrofungus genus Ganoderma. More interestingly, all the enantiomers from Ganoderma fungi, with one exception., are hydroquinone derivatives (602). In addition, the majority of these enantiomers belong to the meroterpenoid class (hydroquinone-terpenoid hybrid), and the terpenyl units here are usually monoterpenoid or sesquiterpenoid. Their structures and names are summarized in Figure 31a,b and Table S31, respectively. The abs. configs. of these enantiomers in this section have all been determined by TDDFT-ECD calculation unless otherwise specified.  Compounds 559-584 represent monomeric hydroquinone-terpenoid enantiomers. Cheng and coworkers discovered a pair of hydroquinone-trinorsesquiterpenoid enantiomers (559a/559b) possessing a fused 6/5/6/6/5 polycyclic skeleton from G. lucidum in 2019 [273]. Compounds 560a/560b−563a/563b, identified as a series of hydroquinonemononorsesquiterpenoid hybrids from G. cochlear in 2014, possess a spiro [4,5]decane ring system (560−562) and an eight-membered ring (563), with the abs. configs. being assigned by single-crystal X-ray diffraction analysis [274]. Compounds 564a/564b from G. lucidum are a pair of rotary door-shaped hydroquinone-normonoterpenoid enantiomers with an unusual 5/5/6/6 ring system, and their abs. configs. were established by interpretation of X-ray crystallographic data [275]. Compounds 565a/565b, a pair of macrocyclic meroterpenoid enantiomers derived from a hydroquinone and an intact sesquiterpenoid, were isolated from G. resinaceum by Chen et al. in 2017 [276]. The hydroquinone-monoterpenoid enantiomers 566a/566b with an unusual dioxacyclopenta[c,d]inden motif were reported from G. applanatum in 2016 [277]. Nine pairs of enantiomers 567a/567b−575a/575b incorporating either monoterpenoid or sesquiterpenoid fragments were obtained from G. applanatum in 2015 [278], and 570a/570b was also reported from G. lucidum in the same year with the abs. configs. being not assigned [279]. Compounds 576a/576b featuring an interesting polycyclic meroterpenoid skeleton with a glycerol unit were isolated from G. applanatum in 2017 [280]. Five pairs of hydroquinone-sesquiterpenoid enantiomers 577a/577b−581a/581b and a racemate 582, all bearing a butenolide fragment, were isolated from G. sinense in 2016 [281]. Compounds 583a/583b and 584a/584b, two pairs of farnesylated hydroquinone enantiomers incorporating a p-hydroxycinnamoyl residue, were discovered from G. sinense in 2016 [281].

Enantiomers from Kingdom Prokaryota
Few enantiomers have been reported from the kingdom Prokaryota, i.e., only five pairs (605-609) to date, all of which were discovered from actinomycetes. Their structures and names are provided in Figure 32 and Table S32 in Supplementary Materials, respectively. Compounds 605a/605b are a pair of angucyclinone enantiomers featuring a unique epoxybenzo[f ]naphtho [1,8-bc]oxocine heterocyclic scaffold, and were isolated from a Streptomyces sp. in 2019, with the abs. configs. being determined by X-ray diffraction analysis [291]. Compound 606, a simple prenylated indole alkaloid bearing a rare cyano group, was isolated as a racemate without further chiral separation from Streptomyces sp. ZZ820 in 2019 [292]. Compounds 607a/607b−609a/609b are three pairs of enantiomeric indole alkaloids with a spiro indolinone-naphthofuran skeleton reported from a Streptomyces sp. in 2017 [293].

Enantiomers from Kingdom Animalia
Compared with those from plants and microorganisms, compounds from animals only account for a small proportion of the large NP family, and have been mainly reported from lower animals such as sponges and corals. Therefore, the number of enantiomers from kingdom Animalia is also limited. According to their biological source, animal-derived enantiomers will be divided into the following three subcategories.

Enantiomers from Phylum Porifera
Animals from phylum Porifera (also termed Spongia) are generally known as sponges. They also represent a very important source of bioactive NPs. Natural enantiomers from sponges mainly include terpenoids and alkaloids; see Figure 33 and Table S34.

Terpenoids
Compounds 610a/610b were identified as a pair of valerenane-type sesquiterpene enantiomers from a Spongia sp. in 2019 [294], while the trinorsesquiterpenoid enantiomers 611a/611b incorporating furan and butenolide rings were isolated from the Beihai sponge Spongia officinalis in 2018, with the abs. configs. being determined by biomimetic total synthesis and modified Mosher's method [295]. The three C 17 norditerpenoid pairs 612−614 with a γ-lactone unit, together with two pairs of sesterterpenoids 615 and 616 with a butenolide unit, were obtained from a Cacospongia sp. in 2019. Among them, all enantiomeric pairs except 614 were successfully separated. Compounds 617a/617b are a pair of sesterterpenoid enantiomers featuring a bicyclo[4.2.0]octene core and were isolated from Hippospongia lachne in 2017 [296]. Compounds 618−621 are four pairs of furanosesterterpene tetronic acids from a Psammocinia sp., and 618 and 619 were found to be geometrical isomers of two pairs of enantiomers as revealed by chiral HPLC analysis. Similar to the case of 618 and 619, compounds 620 and 621 were also proved to be two enantiomeric pairs, but only 620 was finally separated into pure enantiomers [297].

Lipids
Compounds 637a/637b, a pair of interesting C 20 bisacetylenic lipid enantiomers, were discovered from the marine sponge Callyspongia sp. in 2013, with the abs. configs. being determined by modified Mother's method [302]. The lipid zwitterions 638a and 639a were separated from Spirastrella abata in 2012 [303], and their respective enantiomers (638b and 639b) had been previously reported from the same species in 2002 [304].

Enantiomers from Phylum Chordata
There have been only two pairs of enantiomers reported from the animals of phylum Chordata (see Figure 34 and Table S34 in Supplementary Materials). Compounds 647a/647b, a highly nitrogenated enantiomer pair with a novel heterocyclic scaffold incorporating two extra phenol units, were isolated from a marine ascidian Eudistoma sp. in 2016 [308]. A pair of oxygenated myristic acid enantiomers bearing a tetrahydrofuran moiety (648a/648b) was obtained from a larval sea lamprey Petromyzon marinus in 2015, with the abs. configs. being determined by modified Mosher's method [309].

Biological Properties
As is well known, NPs, on one hand, play a decisive role in maintaining their source organisms' health, helping defend against internal or external adverse stresses and enticing favorable stimuli. On the other hand, NPs in the form of herbal medicines have long been used by humans as therapeutic agents against various diseases, thus guaranteeing the continuation of human civilization. With the advances of science and technology, NPs and their derivatives still shine in the research field of modern drug discovery and development [2].
It is widely accepted that chirality, as an important feature of most NPs, is closely related with their bioactivities. Normally, life systems tend to produce/utilize only one molecule of an enantiomeric pair. For example, humans only take in D-glucose and L-amino acids as nutrients. The fact that a pair of enantiomers can exert utterly different bioactivities was recognized as far back as the 1960s, when the 'Phocomelia infants event' caused by the (S)-enantiomer of the synthetic drug thalidomide taught the pharmaceutical industry an important lesson. For many years, however, NP workers failed to recognize the widespread occurrence of enantiomerism in nature, and failed to explore the differences in bioactivity between pairs of enantiomers. Fortunately, as data on natural enantiomers increase in scope, more and more biological properties of different classes of enantiomeric pairs have also been reported, and this has provided more examples with which to investigate the differences in bioactivity among enantiomers.
As bioassay protocols vary in different research labs and even in different batches from the same lab, it should be clarified that we do not intend to invite direct comparisons regarding the activity potency by tabulating the assay data from different reports. Instead, bioactivity comparisons between different labs will be completely avoided in the current review and the use of potency descriptors will also be kept to a minimum. Meanwhile, we will not list the biological data of all reported enantiomers, and only selective cases with obvious activity differences at the enantiomeric level are discussed, under the following subcategories: cytotoxic, antiviral, antibacterial, antifungal, anti-inflammatory, antioxidative, cell protective, enzyme inhibitory, β-amyloid (Aβ) aggregation inhibitory and miscellaneous activities.
The methyl 2-naphthoate enantiomers 390a/390b were found to show inhibition against the proliferation of three types of cancer cells (A549, MCF-7 and MDA-MB-231), with the levoisomer 390a being ca. three times more active than the dextroisomer 390b [8].

Antiviral
The antiviral activities of the enantiomers described in this review are shown in Table 3. The coumarins 254a/254b did not show significant inhibition differences against either the herpes simplex virus 1 (HSV-1) or the host cell between enantiomers, but their racemate (±)-254 exhibited obviously increased activity which was suggestive of a strong synergistic action [66]. Similar synergistic effects of enantiomers were also observed for another two pairs of coumarins, 104a/104b and 105a/105b, with the racemates (±)-104 and (±)-105 displaying 3.2-to 6.1-fold antiviral activity against the influenza virus A (H3N2) compared with the pure enantiomers [67]. Four pairs of phloroglucinol enantiomers 368a/368b−371a/371b were subjected to antiviral assay against Kaposi's sarcoma-associated herpes virus (KSHV); they all showed certain degrees of bioactivity differences at the enantiomeric level [184]. The fungus-derived alkaloid enantiomers 558a/558b and the racemate (±)-558 all exhibited antiviral activity against EV71 virus, with the dextrorotary enantiomer being nearly five times as active as its antipodal isomer [272].

Antibacterial
It appears that most of the antibacterial enantiomeric pairs collected in this review showed remarkably differentiable activities between enantiomers. Nonetheless, a few exceptions were still found and are listed in Table 4. The furoquinoline alkaloid enantiomers 147a/147b were reported to have antibacterial activity against Enterococcus faecalis, and the (-)-enantiomer showed about two-fold activity as the (+)-enantiomer [97]. The phydroxycinnamoylated dihydrochalcone enantiomers 285a/285b−288a/288b exhibited in vitro antibacterial activity against Staphylococcus aureus with IC 50 values ranging from 0.61 to 6.0 µM [153], and it appeared that all the dextrorotary enantiomers were more effective than their respective levorotary isomers, with 288a/288b showing the greatest activity difference, i.e., 3.7 fold [153].

Antifungal
Few reports have been published on the antifungal activities of the enantiomers mentioned in this review, although a handful of examples have shown about two-fold bioactivity differences between enantiomers ( Table 5). The δ-lactone enantiomer (−)-464a was reported to display inhibitory activity against Candida albicans with an MIC of 26.4 µM, while its antipodal enantiomer (+)-464b was considered inactive [219]. In addition, both levorotary enantiomers of compounds (±)-483 and (±)-484 exhibited better antifungal activity again C. albicans than their respective dextrorotary isomers [229], and similar effect against Fusarium solani was also recorded for the indole-piperidine enantiomer pair (±)-524 [256].

Anti-Inflammation
The anti-inflammatory activities of NPs have often been evaluated by testing their inhibitory capability against NO release in LPS-induced BV-2 microglial cells or RAW 264.7 macrophages ( Table 6). The benzofuran-type lignan enantiomers 43a/43b and 44a/44b were tested for their NO production inhibitory effect in LPS-induced BV-2 microglial cells, with (−)-43b and (+)-44a exhibiting pronounced activity with IC 50 values of 8.9 and 5.9 µM, being nearly twice as active as their respective antipodal enantiomers [39]. The levorotary spirodienone lignan enantiomers (−)-82b and (−)-83b showed significant inhibition against NO production in LPS-induced RAW 264.7 macrophages, with both being >3 fold as active as their respective dextrorotary enantiomers [54]. In the same bioassay model, the indolizidine dextroisomer 221a displayed much stronger inhibitory activity (6.3 fold) than the levoisomer 221b [124,128,164]. In contrast, the levorotary enantiomer 407b was much more active (ca. 5 fold) than its antipodal enantiomer 407a in the LPS-induced NO release assay in BV-2 cells [130,132,226,265].

Antioxidation
The DPPH and ABTS radical scavenging assay models have been widely used to evaluate the antioxidative capacity of NPs, although not many of the listed enantiomers in this review have been tested with these bioassays. Owing to their radical mechanism, most tested enantiomers displayed equal potency in both assays as expected, whereas the tryptophan-alanine dipeptide enantiomers 527a/527b-529a/529b showed some activity differences at the enantiomeric level, particularly for dextroisomers 527a and 528a, that showed obviously enhanced radical scavenging activity (4.1 and 2.5 fold, respectively) compared with their levoisomers in the DPPH assay model (Table 7) [257].

Cell Protection
Cell protection assays are usually performed in neuronal cells to explore new chemicals that could be developed for the treatment of neurodegenerative disorders, but which could likely also be used in the search for molecules with which to treat other diseases (Table 8). Generally speaking, a >10% cell viability difference can be considered significant. The protective activity of neolignan enantiomers 24a/24b against H 2 O 2 -induced cell injury in human neuroblastoma SHSY5Y cells was tested and the (+)-enantiomer showed obviously better activity than the (−)-enantiomer [20]. In a same assay model by Zhou et al., the analogous enantiomeric pair 72a/72b also exhibited a similar trend of bioactivity difference, with the dextroisomer displaying better protective effect than the reference drug and the levoisomer being found to be inactive [46]. Further investigations revealed that (+)-72a could significantly decrease the percentages of both early and late apoptotic cells.

Conclusions
As can be seen from Figure 35, the number of identified natural enantiomers steadily increased during the period covered by this study, albeit with slight drops in 2013 and 2018. Notably, more than 100 enantiomers have been reported in the last three years (2017-2019), indicating rapid development in this field. It is also worth noting that plant-derived enantiomers made up 72% of all cases (Figure 36) in the study period, which suggests the continuing vitality of phytochemical studies, despite severe funding cutbacks for traditional NP research in recent years [3]. Another set of statistics ( Figure 37) revealed that alkaloid enantiomers represent the biggest group of molecules from plants, followed by lignans and flavonoids.

Natural Distribution of Enantiomers
As demonstrated by the examples in this review, natural enantiomers have been widely reported from species of all kingdoms except Protoctista, which could be attributed to the fact that few NP researchers have been focusing on Protoctista organisms since they are not well-known sources of interesting molecules. Therefore, the discovery of enantiomers from Protoctista species in the near future is to be expected if NP workers continue to focus on them. From another perspective, the enantiomers collected in the period covered in this review have a broader distribution at the originated species level, from microbial fungi (e.g., mold) to macrofungi (e.g., mushrooms), from lower plants (e.g., moss) to higher plants (e.g., herbs), and from lower animals (e.g., sponges) to higher animals (e.g., fishes). Another noteworthy point is the distribution of enantiomers in different structural families, which can be clearly revealed by the examples in the current and previous reviews [3] that were discovered in all major structural classes such as terpenoids, alkaloids, flavonoids and polyketides (mainly from a biogenetic view). At the lower level of classification, it seems that there have been no enantiomeric cases reported for triterpenoids and steroids. The above-mentioned two points clearly demonstrate the universal occurrence of enantiomerism in nature.

Natural Formation of Enantiomers
It is interesting to note that unlike the previous report [3], in which many enantiomeric examples were obtained from different species, the majority of the cases collected in the current study were isolated from the same species as scalemic or racemic mixtures. Although Williams and colleagues predicted in 2012 [3] that the biogenetic studies of natural enantiomers would be "a fertile area for future inquiry and discovery", there has been no significant progress in this research field since then. Nonetheless, some common reasons or rules regarding enantiomeric production can still be rationalized on the basis of currently available knowledge: (1) For cases in which the two antipodal enantiomers are produced by two different species (from the same or different genus or even different families), such as (+) and (-)-limonenes [3], two distinct enzymes and mechanisms are involved in their biosynthesis; (2) When an enantiomeric pair (racemic or scalemic mixture) is discovered from the same species, the lack (partially or completely) of stereo-specificity of the catalytic enzyme could be responsible for the enantiodivergent formation; (3) The absence of enzyme substrate or a completely chemical process would also lead to the production of two enantiomers, which is especially true for many NPs with only one chiral center. The following two explanations, though not as reasonable as the above-mentioned three, could also not be excluded. (4) In some biochemical processes which involve radicals, though normally stereo-controlled by enzymes, the generation of enantiomers is possible due to the extremely high reactivity of radicals. (5) The extraction and isolation procedures of NPs could also lead to the formation of new chiral centers, and thus, the production of enantiomers [310,311]. At this point, these enantiomeric molecules should be classified as NP derivatives or artifacts.

Structures Tend to Exist as Enantiomers in Nature
With the discovery of more and more enantiomeric NPs containing diverse structures, it can be concluded that enantiomerism may occur for each structural type, although no enantiomers have been reported for triterpenoids and steroids. Compared with enantiomers from plants, those from microorganisms are able to incorporate more complicated structures, e.g., with high molecular weights. It is possible that the enzyme systems in microorganisms are not fully developed and stereoselectivity is lacking. With this investigation into the structures of enantiomers reported from 2012-2019 in hand, we can easily conclude which structures or which groups in the structures tend to exist in nature as enantiomers. (1) NPs contain C6-C3 units in their structures, such as lignans, flavones, coumarins, simple phenylpropanoids, and hybrids between C6-C3 units and other structures. The enantiomerism for those structures presumably derives from the nonstereoselective oxidation of the C3 unit or nonstereoselective coupling of the C6-C3 units, through enzymatic or nonenzymatic reactions. (2) NPs formed by combination of 2~4 isopentenyl units, such as monoterpenoids, sesquiterpenoids, diterpenoids, and meroterpenoids, or having isopentenyl units as side chains, can exist in the form of enantiomers, and need to be further researched. (3) Alkaloid NPs have a variety of structural types, each of which may exist in the form of enantiomers. (4) When NPs with long chain, e.g., fatty acids and diarylheptanoids, have chiral centers, testing whether they are enantiomers or not is necessary. (5) NPs with axial chirality tend to exist as enantiomers in nature.

Identification of the Presence of Enantiomers
The criteria of enantiomeric presence vary, and a confirmative conclusion should be made based on comprehensive considerations. Ideally, the enantiomeric purity of every NP should be checked, but apparently this is neither economical nor technically feasible. Nevertheless, some general guidelines can still be summarized. Firstly, if a NP belongs to a structural group with strong enantiomeric tendency as listed in this review, such as 8,4oxyneolignans, caution is required. Secondly, for a previously undescribed NP, when its [α] D value is very small (e.g., <5) or close to zero, the presence of an enantiomeric mixture should be considered. However, this method is not always fully indicative, as some chiral compounds naturally have low [α] D . For a known NP, regardless of whether the magnitude of [α] D value is big or small, if it obviously deviates from the reported datum, the occurrence of enantiomerism is possible, and the purity of the tested NP should first be guaranteed. Thirdly, ECD measurement can also be used to check the enantiomeric purity of a NP (in case it shows a response in the experiment). A good-quality ECD curve usually looks smooth with clear Cotton effect(s) in the normal wavelength range (mostly 190-400 nm); if not, there is a high probability of enantiomeric presence. The aforementioned empirical knowledge is only based on general cases, and in fact, determination of the presence of enantiomers can be complicated. Notably, when the natural e.e. value of a pair of enantiomers is very high, as in the case of neosecurinane alkaloids [5], the researchers' level of experience and sensitivity to chirality will make the difference.

Separation and Differentiation of Enantiomers
The separation (use of different chiral stationary materials) and differentiation (abs. config. assignment of an enantiomeric pair) of natural enantiomers were well documented in the review by Cass and Batista Jr. [10] and will not be included here. However, we do wish to emphasize that no omnipotent separation material and single technique can be applied for the purification and abs. config. determination, respectively, for all types of enantiomers, and any doubt regarding enantiomeric purity deserves further investigation.

Stereochemistry-Bioactivity Relationship of Enantiomers
As for the stereochemistry-bioactivity relationship (SBR) of natural enantiomers, analyses of the biological data gathered in this review do not provide many meaningful clues, and the relevance between the bioactivity and the chirality (dextroisomer or levoisomer) of a pair of enantiomers seems random and irregular in both enzymatic and cellular level bioassays. Although factors regarding the 'chirality' of life systems are well-known (e.g., Dglucose and L-amino acids as primary metabolites), there is still a long way to go before we are able to reveal the secrets of the exact SBR of enantiomers. Nevertheless, some general conclusions can still be reached according to the presently accessible information, similar to what Prof. Mori described for insect pheromones [4]. For a specific bioassay model: (1) One enantiomer is active, while the opposite enantiomer is less or not active, and the mixture of them does not result in any extra effect; (2) Both enantiomers are equally active, and their mixture does not result in any extra effect; (3) Both enantiomers are inactive or active, but their mixture is active or more active, suggestive of a synergistic action; (4) One enantiomer is active, whereas the antipodal enantiomer exhibits antagonistic activity, and thus, their mixture will exert an offset effect. Please note that the aforementioned general rules vary for different bioassays and are thus to be taken on a case-by-case basis, because all NPs are produced by the source organisms for their own use, and not for use by humans; we simply take advantage of their biological properties.
All in all, notwithstanding the rapidly growing number of reports and improving awareness of natural enantiomers in recent years, there are still a number of questions which remain to be answered. Our understanding of this fascinating natural phenomenon is only in its infancy Here, we would like to say to the NP community that enantiomerism in nature is ubiquitous and vital. We hope that this review will prompt future researchers to routinely ask "Is my natural product enantiomerically pure, and if so, which enantiomer have I obtained?", and in so doing, to perhaps even alter the methods applied by scientists in the future.
Supplementary Materials: The following supporting information can be downloaded at: Tables S1−S34 contain names, source species and references of all collected enantiomers. Refs [312][313][314][315][316] are cited in the Supplementary Materials.