OSMAC-Driven Discovery of Six New Alkaloids from the Cold-Seep-Derived Fungus Talaromyces amestolkiae HDN21-0307

Abstract

Six new alkaloid compounds, including two rare aromatic nitrile compounds talaronitriles A–B (1–2), a novel oxime-functionalized azadiphilone analogue talarooxime A (3), a new phenylhydrazone alkaloid talarohydrazone E (4), and two new dipeptide compounds talarodipeptides A–B (5–6), were isolated from the deep-sea cold-seep-derived fungus Talaromyces amestolkiae HDN21-0307 via OSMAC approach. Compound 1 is the first natural naphthalene compound with cyano groups. Compound 3 represents the first natural product containing an oxime-functionalized azadiphilone scaffold. Their structures and absolute configurations were elucidated through spectroscopic data analysis and quantum chemical calculations. Notably, compound 3 demonstrated moderate DPPH free-radical-scavenging activity, with an IC₅₀ value of 29.41 μM.

1. Introduction

Natural alkaloids are nitrogen-containing alkaline substances found in living organisms and represent one of the most significant classes of secondary metabolites produced by marine fungi [1]. These compounds typically possess structurally complex nitrogenous heterocycles. Owing to their unique chemical scaffolds and diverse biological activities—such as antitumor, antibacterial, antiviral, and anti-inflammatory effects—natural alkaloids have garnered increasing interest in drug research and natural product chemistry [1,2].

Talaromyces fungi are widely distributed across various environments, including marine habitats, soil, and plants. Species within this genus produce an extensive array of secondary metabolites, among which alkaloids represent a major class. To date, numerous alkaloids have been isolated from Talaromyces fungi and can be roughly classified into amide alkaloids, indole alkaloids, terpene alkaloids, and quinoline alkaloids according to their structural types [3,4,5]. Notably, many alkaloids from Talaromyces, such as chaetominine B [3] and ditalaromylectones A–B [4], possess not only novel structures but also promising biological activities, underscoring the remarkable metabolic capacity of Talaromyces fungi for producing novel alkaloids.

In our previous study, the fungal strain Talaromyces amestolkiae HDN21-0307 was isolated from deep-sea cold seep sediments in the South China Sea, yielding four novel phenylhydrazone alkaloids (talarohydrazones A–D) with notable antimicrobial and cytotoxic activities [6]. However, similar to most marine microorganisms, T. amestolkiae HDN21-0307 still harbors numerous biosynthetic gene clusters that remain silent under monoculture conditions, limiting the discovery of novel bioactive metabolites [7]. The OSMAC (One Strain Many Compounds) strategy, which involves altering medium composition and cultivation conditions, has been widely employed to activate these silent gene clusters [8]. To date, this approach remains the most convenient and efficient strategy for mining microbial metabolites [9].

Consequently, the OSMAC strategy was also applied to further explore the metabolites of T. amestolkiae HDN21-0307. Six new alkaloids (1–6) were isolated from T. amestolkiae HDN21-0307 cultivated under two distinct conditions: (1) shaking conditions in Modified Glucose Minerals Salts 2# liquid medium and (2) static conditions in rice solid media with the addition of NaBr. Notably, these included the first reported naphthalene-fused nitrile natural product (1) and the inaugural oxime-functionalized azaphenanthrenone alkaloid (3). Additionally, compound 3 showed moderate DPPH free-radical-scavenging activity, with an IC₅₀ value of 29.41 μM. Herein, details of the isolation, structure elucidation, and bioactivity evaluation of these compounds are presented.

2. Results

The OSMAC strategy was applied to comprehensively explore the metabolic potential of T. amestolkiae HDN21-0307 by employing six different culture media under shaking or static cultivation conditions. As a result, this strain exhibited distinct UV absorption profiles compared with those of previously known compounds from T. amestolkiae HDN21-0307 [6] under the following conditions: (1) modified Glucose Mineral Salts 2# liquid medium under shaking conditions and (2) rice solid medium supplemented with NaBr under static conditions (Figure S1). To elucidate the structures corresponding to these characteristic peaks, large-scale fermentations were conducted under both conditions, followed by ethyl acetate and methanol extraction. The resulting extracts were subjected to extensive column chromatography over silica gel, ODS, and Sephadex LH-20, ultimately yielding six new alkaloids (1–6) (Figure 1).

Compound (1), obtained as a colorless powder, has a molecular formula of C₂₀H₁₄N₂O₂ based on HRESIMS (m/z 315.1128, [M + H]⁺ calcd. for C₂₀H₁₅N₂O₂ 315.1124), indicating fifteen degrees of unsaturation. Analysis of 1D NMR and HSQC (

Table 1. ¹H NMR (400 MHz) and ¹³C NMR (125 MHz) data of 1 in methanol-d₄ and 2 in chloroform-d (δ in ppm, J in Hz).

No.	1		No.	2
	δC, Type	δH (J in Hz)		δC, Type	δH (J in Hz)
1	148.2, C		1	119.1, C
2	111.7, C		2	106.1, C
3	108.5, C		3	100.8, C
4	136.2, CH	8.50, s	4	116.1, C
4a	130.5, C		1′	125.4, C
5	132.2, C	8.07, d (9.0)	2′	132.5, CH	7.41, d (8.9)
6	123.8, CH	7.46, dd (9.0, 2.5)	3′	114.4, CH	6.93, d (8.9)
7	162.9, CH		4′	161.9, C
8	107.1, CH	7.06, d (2.5)	5′	114.4, CH	6.93, d (8.9)
8a	136.5, C		6′	132.5, CH	7.41, d (8.9)
9	117.0, C		7′	146.1, CH	7.35, s
10	117.7, C		8′	55.6, CH3	3.88, s
11	56.0, CH3	3.77, s	1″	125.4, C
1′	128.9, C		2″	132.1, CH	7.87, d (8.9)
2′	132.3, CH	7.42, d (8.7)	3″	114.8, CH	6.98, d (8.9)
3′	115.5, CH	7.18, d (8.7)	4″	162.7, C
4′	162.2, C		5″	114.8, CH	6.98, d (8.9)
5′	115.5, CH	7.18, d (8.7)	6″	132.1, CH	7.87, d (8.9)
6′	132.3, CH	7.42, d (8.7)	7″	149.0, CH	7.46, s
7′	55.9, CH3	3.93, s	8″	55.7, CH3	3.85, s

) indicated the existence of eighteen aromatic carbons, including ten non-hydrogenated carbons (δ_C 108.5, 111.7, 117.0, 117.7, 128.9, 130.5, 136.5, 148.2, 162.2, and 162.5), and two oxymethyl carbons (δ_C 55.9 and 56.0). The characteristic hydrogen signals H-2′/H-6′ (δ_H 7.42, d, J = 8.7 Hz) and H-3′/H-5′ (δ_H 7.18, d, J = 8.7 Hz) indicated the presence of a para-substituted benzene ring. Among them, the substituent at the 4′-position of the benzene ring was unambiguously identified as a methoxy group (C-7′), as confirmed by its characteristic chemical shift (C-4′, δ_C 162.2; C-7′, δ_C 55.9) and key HMBC correlation from H-7′ (δ_H 3.93, s) to C-4′. Additionally, a trisubstituted benzene ring with C-7 oxygen methyl group was determined through characteristic signals of H-5 (δ_H 8.07, d, J = 9.0 Hz), H-6 (δ_H 7.46, dd, J = 9.0, 2.5 Hz), and H-8 (δ_H 7.06, d, J = 2.5 Hz) and the HMBC correlations from H-11 (δ_H 3.77, s) to C-7 (δ_C 162.9), from H-5 to C-7/C-8a (δ_C 163.0), from H-6 to C-4a (δ_C 130.5), and from H-8 to C-4a/C-6 (δ_C 123.8)/C-7.

The established structural fragments account for eight degrees of unsaturation, leaving seven remaining unsaturation to be assigned. Analysis of the NMR data indicated only six aromatic carbons (δ_C 108.5, 111.7, 117.0, 117.7, 136.2, and 148.2) and two nitrogen atoms remained unassigned. To satisfy the high degree of unsaturation, compound 1 likely contained an additional aromatic ring. The key HMBC correlations of H-8/C-1, H-5/C-4 (δ_C 163.2), and H-4/C-2/C-5/C-8a confirmed the pentasubstituted aromatic ring as part of a poly-substituted naphthalene scaffold. Furthermore, the remaining two non-protonated aromatic carbons (C-9, δ_C 117.0; C-10, δ_C 117.7) and two nitrogen atoms needed to account for four additional degrees of unsaturation, leading to the identification of two cyano groups in compound 1. Finally, the para-substituted benzene ring was determined to be connected to the C-1 of the naphthalene scaffold through the correlations between H-2′/H-6′ and C-1 (δ_C 148.2), while the remaining two cyanide groups were connected to C-2 (δ_C 111.7) and C-3 (δ_C 108.5), respectively. In addition, the presence of a cyano group was also confirmed by its characteristic infrared absorption band at 2227 cm⁻¹. The structure of 1 was conclusively established and named talaronitrile A [10].

The cyano group is widely used in the pharmaceutical industry for its ability to enhance binding affinity, improve pharmacokinetic properties, and reduce drug resistance [11,12]. Naturally occurring nitriles, mostly existing as cyanogenic glycosides, are widely distributed across various organisms including plants, fungi, and marine species [11,12]. Aromatic nitrile-containing natural products are relatively rare [11,12]. Notably, compound 1 represents the first natural product featuring a naphthalene-linked nitrile scaffold.

Compound (2) was obtained as a pale-yellow powder with a molecular formula of C₂₀H₁₆N₂O₂, based on HRESIMS data (m/z 317.1294 [M + H]⁺ calcd. for C₂₀H₁₇N₂O₂, 317.1285). Comparative analysis of the 1D NMR and HSQC spectra (Table 1) revealed the existence of sixteen aromatic carbons, including eight non-hydrogenated carbons (δ_C 100.8, 106.1, 116.1, 119.1, 125.4, 125.4, 161.9, and 162.7) and two oxymethyl carbons (δ_C 55.6 and 55.7). Compound 2 and epurpurin C shared the similar nitrile skeleton, as evidenced by the HMBC correlations from H-7′ (δ_H 7.35, s) to C-1 (δ_C 119.1)/C-3/C-2′, from H-2′/6′ (δ_H 7.41, d, J = 8.9 Hz) to C-1′/C-4′ (δ_C 161.9)/C-7′, from H-3′/5′ (δ_H 6.93, d, J = 8.9 Hz) to C-1′/C-4′, from H-7″ (δ_H 7.46, s) to C-2/C-4 (δ_C 116.1)/C-2″, from H-2″/H-6″ (δ_H 7.87, d, J = 8.9 Hz) to C-1″/C-4″ (δ_C 162.7)/C-7″, and from H-3″/5″ (δ_H 6.98, d, J = 8.9 Hz) to C-1″/C-4″ (Figure 2) [13]. The only difference between 2 and epurpurin C was that the absolute configuration of Δ^3,7″ in 2 was E, supported by the existence of asymmetrical peaks in 1D NMR. Additionally, the C-H coupling constants (³J_C-4,H-7″ = 14.6 Hz) further corroborated the double-bond configurations of Δ^3,7″ (Figure S18) [14]. Hence, the structure of 2 was established and named talaronitrile B.

Compound (3), obtained as a bright yellow powder, has a molecular formula of C₂₁H₁₇NO₁₀, based on HRESIMS data (m/z 444.0927 [M + H]⁺ calcd. for C₂₁H₁₈NO₁₀, 444.0925). The 1D NMR and HSQC spectra (

Table 2. ¹H NMR (500 MHz) and ¹³C NMR (125 MHz) data of 3 in acetone-d₆ and 4 in DMSO-d₆ (δ in ppm, J in Hz).

No.	3		No.	4
	δC, Type	δH (J in Hz)		δC, Type	δH (J in Hz)
1	189.0, C		1
2	86.9, C		2	160.9, C
3	185.4, C		3	127.9, C
4	142.4, C		4	178.2, C
4a	120.0, C		5	108.6, CH	5.91, s
5	104.9, CH	6.56, s	6	145.1, C
6	158.8, C		7	180.5, C
8	64.6, CH2	5.14, m	8	134.6, C
8a	141.3, C		9	129.9, CH	7.91, m a
9	137.3, CH	7.23, d (15.5)	10	129.0, CH	7.61, m a
10	124.4, CH	6.42, d (15.5)	11	134.4, CH	7.76, m a
11	166.9, CH		12	128.6, CH	7.61, m a
12	21.9, CH3	1.77, s	13	129.1, CH	7.91, m a
1″	170.5, C		16	122.6, C
2″	104.3, C		17	151.6, C
3″	166.6, C		18	99.7, CH	6.62, d (2.3)
4″	101.8, CH	6.25, d (2.4)	19	159.8, C
5″	164.4, C		20	109.3, CH	6.57, dd (8.8, 2.1)
6″	113.0, CH	6.38, d (2.1)	21	117.9, CH	7.69, d (8.3)
7″	145.0, C		22	56.3, CH3	3.92, s
8″	24.3, CH3	2.62, s	1	-NH	11.15, s
3″	-OH	10.68, s	15	-NH	17.30, s
			19	-OH	10.30, s

^a Signals were overlapped.

) revealed the presence of seventeen aromatic carbons, including thirteen non-hydrogenated carbons (δ_C 86.9, 104.3, 120.0, 141.3, 142.4, 145.0, 158.8, 164.4, 166.6, 166.9, 170.5, 185.4, and 189.0), five methine carbons (δ_C 101.8, 104.9, 113.0, 124.4, and 137.3), one methylene carbon (δ_C 64.6), and two methyl carbons (δ_C 21.9, 24.3). The ³J_H-9,H-10 (15.5 Hz) unambiguously assigned the E configuration of the double bond Δ^9,10. Comparative NMR data revealed that compound 3 and talarohydrazone B share an essentially identical core structure (Figure 2). The key structural differences include the absence of a phenylhydrazone moiety in compound 3, along with the presence of an additional nitrogen atom and hydroxyl group. The remaining degree of unsaturation, combined with the characteristic downfield shift of C-4 (δ_C 142.4), suggested the presence of an oxime moiety at C-4 in 3. Consequently, the planar structure of 3 was established and designated as talarooxime A.

Due to the lack of diagnostic NOESY correlations, the configuration of the Δ^4,N−13 double bond could not be determined spectroscopically. To resolve this, the time-dependent density-functional theory (TDDFT) calculations were performed on both possible diastereomers 4E-3 (3a) and 4Z-3 (3b) (Figure S55). DP4+ probability analysis demonstrated excellent agreement (99.94%) between 3b and the experimental NMR data, establishing the Z-configuration at Δ^4,N−13 double bond (Figure S56). The C-2 configuration was assigned as consistent with talarohydrazone B based on biosynthetic considerations. Furthermore, (2S)-3 showed remarkable concordance between experimental and calculated ECD spectra (Figure 3), thereby confirming the absolute configuration at C-2. Thus, the absolute configuration of compound 3 was determined to be (2S, 4Z, 9E). Structural diversity among currently known azaphilones is predominantly concentrated on chain extension at the C-2/C-6 positions of the compound 3 core scaffold, with modifications at the C-4 position being rare. Furthermore, the substituents introduced are predominantly aliphatic chains or aryl/heteroaryl groups [15]. In contrast, compound 3 represents the first reported oxime-containing azaphilone derivative featuring a substitution at the C-4 position.

Compound (4) was isolated as a red amorphous powder. The molecular formula of 4 was determined to be C₁₉H₁₅N₃O₅ based on the HRESIMS date (m/z 364.0937 [M − H]⁻ calcd. for C₁₉H₁₄N₃O₅, 364.0939), indicating fourteen degrees of unsaturation. Spectroscopic analysis of the ¹H NMR data of 4 showed signals for three active hydrogen protons (δ_H 10.30, 11.15, and 17.30), nine olefinic protons, and one methoxy group. Analysis of the ¹³C NMR and HSQC spectra (Table 2) indicated three carbonyl carbons (δ_C 161.5, 178.9, and 180.5), fifteen olefinic carbons, including six non-hydrogenated carbons, and one oxymethyl (δ_C 56.1) carbon. The aforementioned information accounted for eleven out of the fourteen degrees of unsaturation, suggesting that 4 was a tricyclic compound. The NMR data of 4 was highly similar to that of talarohydrazone A (Figure 2) [6], with the only difference being the presence of a carbonyl group at C-7 in 4 instead of a methylene group at C-7 in talarohydrazone A. The HMBC correlations from H-5/H-9/H-13 to C-7 (δ_C 180.5) confirmed this difference between them. Thus, the planar structure of 4 was determined, named talarohydrazone E. The configuration of the double bond Δ^3,14 could not be determined due to the absence of NOESY correlations. TDDFT was used to calculate NMR data of the two possible configurations 3E-4 (4a) and 3Z-4 (4b). The improved DP4+ probability analysis showed that 4a was in agreement with the NMR experimental data with a probability of 100% (Figure S58 and Table S3), confirming the configuration of the double bond Δ^3,14 as E configuration.

Compound (5) was purified as a pale-yellow oil with the molecular formula as C₁₅H₁₄N₂O₂ based on the HRESIMS data (m/z 255.1126, [M + H]⁺ calcd. for C₁₅H₁₅N₂O₂, 255.1128). The ¹³C NMR and HSQC spectra of 5 indicated the presence of fifteen carbon atoms (

Table 3. ¹H NMR (400 MHz) and ¹³C NMR (100 MHz) data of 5 in chloroform-d and 6 in methanol-d₄ (δ in ppm, J in Hz).

No.	5		No.	6
	δC, Type	δH (J in Hz)		δC, Type	δH (J in Hz)
1			1	167.1, C
2	156.8, C		2	134.1, C
3	133.5, C		3
4	120.3, CH	6.26, t (3.1)	4	164.8, C
5	28.3, CH2	2.84, td (9.1, 3.1)	5	126.1, C
6	46.0, CH2	4.13, t (9.1)	6	114.7, CH	6.53, dd (3.8, 1.1)
7			7	110.5, CH	6.03, dd (3.8, 2.6)
8	156.79, C		8	123.6, CH	6.87, dd (2.4, 1.2)
9	131.7, C		9
10	56.3, CH3	2.91, s	10	35.9, CH3	3.06, s
11	119.6, CH	7.30, s	11	137.9, CH	7.63, s
12			12	53.2, CH3	3.68, s
1′	134.7, C		1′	134.0, C
2′	129.4, CH	7.22, t (7.5)	2′	131.2, CH	7.61, ma
3′	128.3, CH	7.33, ma	3′	130.2, CH	7.40, m a
4′	128.2, CH	7.33, ma	4′	131.9, CH	7.40, m a
5′	128.3, CH	7.33, ma	5′	130.2, CH	7.40, m a
6′	128.3, CH	7.22, t (7.5)	6′	131.2, CH	7.61, m a

^a Signals were overlapped.

), including two carbonyl carbons (δ_C 156.79 and 156.84,), ten olefinic carbons, two methylenes (δ_C 28.3 and 46.0,), and one methyl (δ_C 36.3). The NMR data of 5 closely resembled that of cyclo Pro-1-N-methyl-9,11-en-Phe [16], indicating that they possessed the same cyclic dipeptide skeleton. The only structural difference between them lies in the bond between C-3 and C-4: a double bond in 5 versus a single bond in cyclo Pro-1-N-methyl-9,11-en-Phe. Based on the COSY correlations of H-4 (δ_H 6.26)/H-5, and the HMBC correlations from H-5 to C-3 (δ_C 133.5) and from H-6 to C-4 (δ_C 120.3), the double bond Δ³ in 5 was established (Figure 2). Thus, the planar structure of 5 could be determined, named talarodipeptide A. Due to the absence of NOESY correlations of H-10/H-11, the configuration of the double bond Δ^9,11 was determined by HSQMBC (Figure S44). The C-8 to H-11 (J_C,H = 9.6 Hz) obtained from the HSQMBC spectrum unambiguously assigned the E configuration of the double bond Δ^9,11 [14].

Compound (6) was isolated as a pale-yellow oil. The molecular formula, C₁₆H₁₆N₂O₃, was deduced from its HRESIMS data (m/z 285.1237, [M + H]⁺ calcd. for C₁₆H₁₇N₂O₃, 285.1234), indicating 10 degrees of unsaturation. The ¹³C NMR and HSQC spectra of 6 indicated (Table 3) two carbonyl carbons (δ_C 164.8 and 167.1), twelve olefinic carbons, including three non-hydrogenated carbons, and two methyls (δ_C 35.9 and 53.2,). The NMR data of 6 closely resembled that of Pyrrole-Phe-OMe ester [17], indicating that they had the same skeleton. The difference was that the 3-NH of 6 was replaced by a methyl group (C-3, δ_C 35.9), and C-2 (δ_C 134.1) and C-11 (δ_C 137.9) were double-bonded rather than single-bonded, confirmed by the HMBC correlations from H-10 (δ_H 3.06, s) to C-2/C-4, and from H-2′ to C-11 (Figure 2). The NOESY correlation of H-10/H-11 indicated the E-configuration of the double bond Δ^2,11. Thus, the structure of 6 was determined, named talarodipeptide B.

Antibacterial, antitumor, and the free-radical-scavenging activity tests were conducted on compounds 1–6. Only compound 3 exhibited strong activity (IC₅₀ = 29.41 μM) in the free-radical scavenging. Vitamin C (IC₅₀ = 14.8 μM) was used as the positive control.

3. Materials and Methods

3.1. General Experimental Procedures

Ultraviolet (UV) spectroscopy was recorded on a Hitachi 5430 spectrophotometer (Hitachi, Tokyo, Japan). Circular dichroism (CD) spectra and optical rotations were measured by JASCO P-1020 digital polarimeter (JASCO Corporation, Tokyo, Japan), respectively. Infrared (IR) spectra were obtained on a Bruker Tensor-27 spectrophotometer (Bruker, Saarbruecken, Germany) with KBr pellets. High-resolution electrospray ionization mass spectrometry (HRESIMS) data were acquired on a Thermo Scientific LTQ Orbitrap XL mass spectrometer (Thermo Fisher Scientific, Waltham, MA, USA). Nuclear magnetic resonance (NMR) spectra were recorded on a JEOL JNM-ECP 600 MHz spectrometer (JEOL, Beijing, China) and an Agilent 500 MHz DD2 spectrometer (Agilent, Beijing, China), with tetramethylsilane (TMS) as an internal standard. Column chromatography was performed using Sephadex LH-20 (Amersham Biosciences, Uppsala, Sweden) and silica gel (Qingdao Marine Chemical Factory, Qingdao, China). Preparative high-performance liquid chromatography (HPLC) was carried out on a YMC-Pack ODS-A column (10 × 250 mm, 5 μm, 3 mL/min; YMC Co., Ltd., Kyoto, Japan).

3.2. Fungal Material and Fermentation

T. amestolkiae HDN21-0307 (GenBank No. OQ954837) was isolated from deep-sea cold seep sediments in the South China Sea (119°17′005.887″ E, 22°07′02.589″ N). This strain is a facultative anaerobe. Its mycelium, initially off-white, developed a green pigmentation after 4–5 days of incubation. Additionally, the strain produced a diffusible red pigment over time. This strain has been deposited in the Marine Medicinal Bioresources Center, School of Medicine and Pharmacy, Ocean University of China, Qingdao, China.

3.3. OSMAC Research, Fermentation, and Extraction

Based on the OSMAC strategy, T. amestolkiae HDN21-0307 was cultured in two different ways—shaking condition and static condition.

Shaking conditions: Using five distinct liquid media formulations: Modified Fungus 2# liquid media (Table S1. N1), PDB liquid media (Table S1. N2), SDA liquid media (Table S1. N3), Modified Glucose Minerals Salts 1# liquid media (Table S1. N4), and Modified Glucose Minerals Salts 2# liquid media (Table S1. N5). The products of T. amestolkiae HDN21-0307 cultured in MGMS (Table S1. N5) liquid media 2# under shaking conditions displayed unusual UV absorptions, which was the reason for selecting as the optimal conditions for large-scale fermentation. T. amestolkiae HDN21-0307 was cultured in 500 mL Erlenmeyer flasks, containing 150 mL of GMS liquid media for 9 days under shaking conditions (180 rpm, 28 °C, pH 7.0) in seawater (collected from JiaoZhou Bay, Qingdao, China). A total of 45 L of broth was extracted using EtOAc (3 × 45 L) to obtain crude extract (45.0 g).

Static conditions: Using two distinct solid media formulations: Rice solid media (Table S1. N6), NaBr rice solid media (Table S1. N7). The HPLC analysis of extracts revealed the chromatogram of the extract from the strain HDN21-0307 cultured in rice media with the addition of NaBr exhibited different peaks compared to rice media with no addition. Therefore, rice media with the addition of NaBr was selected as the optimal condition for large-scale fermentation. The fungus was cultured in 500 mL Erlenmeyer flasks each containing 80 g rice, 1 g NaBr and 120 mL freshwater at 28 °C for 30 days. A total of 12 L of broth was extracted using methanol (3 × 12 L) to obtain crude extract (100.0 g).

3.4. Separation and Purification of Compounds

The shaking crude extract was fractionated into five subcomponents (Fr.A-Fr.E) through VLC column chromatography, utilizing a stepped gradient elution protocol with MeOH-CH₂Cl₂. Fr.C was separated by ODS column to obtain five subfractions (Fr.C.1-Fr.C.5). Fr. C.2 was purified by Sephadex LH-20 and semi-preparative HPLC (YMC-pack ODS, 10 × 250 mm, 3.0 mL/min) to afford 3 (3.2 mg, t_R = 25 min). Fr.C.4 was purified by semi-preparative HPLC to afford 1 (3.2 mg, t_R = 31 min). Fr.C.4 was purified by semi-preparative HPLC to afford 2 (2.1 mg, t_R = 29 min).

The static crude extract also used the same steps to separate the extract into the five subfractions (Fr.A-E). Fr. C was separated by ODS column to obtain four subfractions (Fr.C.1-Fr.C.4). Fr. C.1 was purified by semi-preparative HPLC to afford 5 (5.0 mg, t_R = 24 min). Fr.C.2 was purified by semi-preparative HPLC to afford 6 (6.0 mg, t_R = 24 min). Fr.C.3 was purified by semi-preparative HPLC to afford 4 (2.0 mg, t_R = 25 min).

Talaronitrile A (1): white powder, UV (MeOH) λ_max 223 (0.3), 258 (1.0) nm; IR (KBr) ν_max 2925, 2227, 1677, 1614, 1206 cm⁻¹; ¹H and ¹³C NMR data, Table 1; HRESIMS m/z 315.1124 [M + H]⁺ (calcd. for C₂₀H₁₅N₂O₂, 315.1128).

Talaronitrile B (2): faint yellow powder, UV (MeOH) λ_max 241 (0.3), 311 (0.4) nm, 360 (0.5) nm; IR (KBr) ν_max 3419, 2921, 2221, 1602, 1510, 1263, 1178, cm⁻¹; ¹H and ¹³C NMR data, Table 1; HRESIMS m/z 317.1294 [M + H]⁺ (calcd. for C₂₀H₁₇N₂O₂, 317.1285).

Talarooxime A (3): green yellow powder, ${\left[\alpha \right]}_{\mathrm{D}}^{25}$ 150 (MeOH); UV (MeOH) λ_max 215 (1.0), 265 (1.0) nm, 406 (0.4) nm; IR (KBr) ν_max 1646, 1536, 1324, 1267, 1172 cm-1; ¹H and ¹³C NMR data, Table 2; HRESIMS m/z 444.0927 [M + H]⁺ (calcd. for C₂₁H₁₈NO₁₀, 444.0925).

Talarohydrazone E (4): red amorphous powder; UV (MeOH) λ_max 206 (1.86), 258 (0.81), 510 (1.39) nm; IR (KBr) ν_max 3446, 2923, 1626, 1417, 1307, 1284, 1024 cm⁻¹; ¹H and ¹³C NMR data, Table 2; HRESIMS m/z 364.0937 [M − H]⁻ (calcd. for C₁₉H₁₄N₃O₅, 364.0939).

Talarodipeptide A (5): pale-yellow oil; UV (MeOH) λ_max 225 (0.52), 308 (1.25) nm; IR (KBr) ν_max 3436, 1682, 1435, 1136, 760 cm⁻¹; ¹H and ¹³C NMR data, Table 3; HRESIMS m/z 255.1126 [M + H]⁺ (calcd. for C₁₅H₁₅N₂O₂, 255.1128).

Talarodipeptide B (6): pale-yellow oil; UV (MeOH) λ_max 220 (0.65), 274 (1.02) nm; IR (KBr) ν_max 3279, 1717, 1612, 1544, 1437, 762 cm⁻¹; ¹H and ¹³C NMR data, Table 3; HRESIMS m/z 285.1237 [M + H]⁺ (calcd. for C₁₆H₁₇N₂O₃, 285.1234).

3.5. Computation Section

The details of ECD calculation for 3 and the NMR calculation for 3–4 are available in the Supporting Information (Figures S54–S58 and Tables S2 and S3).

3.6. Assay of Activity

DPPH activity. The free-radical clearance experiment is based on the Sharma method [18] and has been modified. Compound 1–3 and vitamin C are dissolved in anhydrous ethanol and diluted into 6 gradients. Mix the vortex of these samples evenly, avoid light at room temperature for 30 min, and read the absorption at 515 nm as a positive control.

The ability to scavenge the DPPH was calculated according to the equation:

DPPH free-radical-scavenging rate D_VC% = [(A_blank − A_{postive control}) ÷ A_blank] × 100%.

DPPH free-radical-scavenging rate D_sample% = [[A_blank − (A_sample − A_control)] ÷ A_blank] × 100%.

A_blank: the absorbance of the DPPH solution;

A_{postive control}: the absorbance of the solution with VC;

A_control: absorbance of the sample.

Antibacterial activity. Antibacterial activities of compounds 1–6 were evaluated against Bacillus cereus ATCC 4342, MRCNS, MRSA, Staphylococcus aureus ATCC 29213, Klebsiella pneumoniae ATCC 43816, Escherichia coli ATCC 25922, Enterococcus faecalis ATCC 29212, Pseudomonas aeruginosa ATCC 27853, and Bacillus subtilis ATCC 6051. Ciprofloxacin was used as a positive control. The detailed methodologies for biological testing have been described in a previous report [19].

Cytotoxicity Assay. Cytotoxic activities of compounds 1, 4–6 were evaluated against K562 (using the MTT method), MDA-MB-231, NCI-H446, NCI-H446/EP, and ASPC-1 (using the SRB method) cell lines. Adriamycin was used as a positive control. The detailed methodologies for biological testing have been described in previous reports [20,21].

4. Conclusions

In summary, the application of the OSMAC strategy successfully activated the secondary metabolic potential of T. amestolkiae HDN21-0307, leading to the isolation of six new alkaloids. Among these, compounds (1) and (2), isolated for the first time from the genus Talaromyces, represent cyano-containing natural products. Notably, compound (1) constitutes the first naturally occurring cyano-substituted naphthalene derivative, while compound (3) is the first natural product featuring an azaphilone scaffold bearing an oxime functional group. This result emphasizes the significance of the OSMAC approach in unlocking the metabolic potential of microorganisms and further demonstrates the substantial potential for discovery within the secondary metabolites of fungi originating from extreme habitats such as cold springs.