Hermetia illucens for Replacing Fishmeal in Aquafeeds: Effects on Fish Growth Performance, Intestinal Morphology, and Gene Expression in the Zebrafish (Danio rerio) Model

Abstract

For improving aquafeed sustainability, insect meal is currently considered the most promising alternative to fishmeal. However, in this regard, more data are still necessary to avoid possible negative impacts on fish growth performance, metabolism, and welfare. The present study investigated the effects of increasing the inclusion of Hermetia illucens meal (0%, 17%, 33% and 50% of the feed, equating to 0%, 34%, 66% and 100% fishmeal replacement) on fish mortality, growth performance, intestine morphology, and gene expression of intestinal carriers. The results showed no adverse effects on fish mortality, feed intake and body weight and a positive effect on feed conversion ratio. Body weight gain was higher when 17% and 50% of Black soldier fly meals’ feed included (34% and 100% fishmeal replacement, respectively). Gut morphology was not affected by the dietary treatments except for the area of PAS-positive goblet cells that was higher in the treatment fed 33% of insect meal. The mRNA expression of intestinal epithelium functionality-specific marker genes, such as slc15a1 (alias pept1, alias slc15a1b), gata4 and nfkb1b, confirmed that the insect meal-based diets might replace fishmeal-based diets without negative effects. Overall, the results of the present study suggest that using Hermetia illucens larvae meal as a replacement for fishmeal in aquafeeds might help to enhance sustainability while assuring favorable fish growth performance and gut health.

1. Introduction

Insect protein meal is currently considered a viable alternative to meals made from fish (fishmeal) and other aquatic sources due to its nutrient profile, digestibility, and palatability. This view is confirmed and supported by several studies demonstrating positive growth and health performance in fish receiving insect meal [1,2,3,4].

The use of insect meal also implies relevant advantages in terms of environmental sustainability and, in perspective, economic viability [5]. Farming insects might be highly convenient thanks to their favorable reproduction and growth rate, high feed conversion efficiency and because they can be reared on a wide array of bio-wastes. Insect production requires low energy input resulting in a low emission of greenhouse gases and ammonia [2,6,7]. Nutrient composition of insect meal may significantly vary according to the species used, the nutritional composition of the rearing substrates and the production processes [5]. Upon standardized and fully controlled production process, insect meal represents a very interesting protein source to be used for feeding many farmed fish species [1,3,8,9]. For the above-mentioned reasons, since 2017, its uses in aquafeeds have been approved in the EU (Commission Regulation 2017/893).

Among the five major groups of insects so far investigated (i.e., Black soldier fly, House fly, Mealworm beetles, Locusts-grasshoppers-crickets, and Silkworm), the Black soldier fly (Hermetia illucens) has been defined as one of the most suitable species for insect meal production [10]. Black soldier fly larvae meal is a high-value raw material rich in fat and protein [2]. The crude protein (CP) can range from 30 to 60% [11,12], and it has been demonstrated to be a suitable replacement for soybean meal in pigs and poultry diets [2,13,14]. Several studies have also shown that Hermetia illucens meal can partially or fully replace fishmeal in diets for several fish species without a negative impact on fish growth performances and health [3,4,12,15,16,17,18,19,20]. However, in some cases, a limited trend (not significant) to reduce growth performance was reported in Rainbow trout (Oncorhynchus mykiss) and Atlantic salmon (Salmo salar) [15,16]. In turbot (Psetta maxima), a significant reduction of feed conversion and growth rates were observed when insect meal inclusion was higher than 33% [17]. Some studies also reported that the dietary inclusion of Black soldier flies larvae meal did not alter the sensory quality of fillets [16,18].

In several terrestrial animals, such as poultry and pigs, diets significantly modified morphology, histology, and biochemical characteristics of the gastrointestinal tract [21,22]. The effect of insect meal inclusion on fish gastrointestinal tracts is still debated and not yet fully understood. In Jian carp (Cyprinus carpio var. Jian), the dietary inclusion of Hermetia illucens meal higher than 8% (75% fishmeal replacement rate) resulted in pathological alteration of intestinal tissue, a higher content of debris within intestinal microvilli, and an increase of HSP70 relative gene expression in hepatopancreas [22]. Dumas et al. [23] reported that shorter villi in Rainbow trout (Oncorhynchus mykiss) fed a diet containing 26.4% of Hermetia illucens meal, while no differences occurred using lower inclusion levels (6.6% and 13.2%). Conversely, Renna et al. [24] demonstrated that a 40% inclusion rate of Hermetia illucens meal in the Rainbow trout (Oncorhynchus mykiss) diet did not affect villus length.

Consistently with recent studies, more effects have been reported on gene expression/modulation, including many gut-related genes [25,26,27,28].

The objective of this study was to investigate whether the dietary inclusion of Hermetia illucens meal and fishmeal replacement rates higher than those previously investigated [4] can determine the potential negative impact on fish growth performances as also suggested, for instance, by Kroeckel et al. [17]. In detail, the effects of the dietary inclusion of 17%, 33% and 50% of insect meal (25%, 50% and 100% fishmeal replacement rate, respectively) on survival rate, growth performance, feed intake, feed conversion rate, intestine morphology, and gene expression, were investigated. To this purpose, adult Zebrafish (Danio rerio) were used as fish model.

2. Materials and Methods

2.1. Animals and Breeding System

The study was carried out at the Zebrafish facility of the Department of Veterinary Sciences, University of Pisa, according to the EU and National regulations for animal welfare (Prot. n. 0039681/2017).

The breeding system and water parameters were those suggested for Zebrafish by Westerfield [29]. A “stand-alone” and water recirculating system hosted the 16 tanks (Tecniplast^® S.p.A., Buguggiate, Italy), 3.5 L capacity each, used for the experiment. Water flow was set at 2 L/h, photoperiod 12/12 dark/light, and water parameters maintained in the following ranges: temperature 28 ± 0.5 °C, pH 7.2–7.8; electrical conductivity between 600 and 800 µS/cm; dissolved oxygen above 5 mg/L; ammonia, nitrites, and nitrates maintained below 1, 0.25, and 50 mg/L, respectively.

2.2. Experimental Design

A total of 320 Zebrafish (wild-type, AB strain; BW 43.6 ± 22.96 mg), 60 days post-fertilization (dpf), were distributed into 16 tanks (20 fish per tank; 5.7 fishes/L). Four experimental diets were formulated with an increasing insect meal (IM) inclusion level and fishmeal (FM) replacement rate, as follows: IM0 (control, 0% IM and 50% FM; 0% FM replacement), IM17 (17% IM and 33% FM; 34% FM replacement), IM33 (33% IM and 17% FM; 66% FM replacement) and IM50 (50% IM and 0% FM; 100% FM replacement). For each treatment, four replicates were used.

The IM obtained from Black soldier fly (Hermetia illucens) larvae (Insect Protein Meal^®, Protix^®, Dongen, The Netherlands) was used as a protein source and as an alternative to FM. The formulated extruded feeds (400–600 µm particle size) were produced by Sparos^® (Olhão, Portugal), and the ingredients and proximal composition are in

Table 1. Ingredients and proximate composition of the experimental diets.

Ingredient	IM0	IM17	IM33	IM50
	%	%	%	%
Fishmeal 1	50.00	33.00	17.00	-
Black soldier fly larvae meal 2	-	17.00	33.00	50.00
Squid meal	7.00	7.00	8.00	9.00
Wheat meal	7.00	6.00	5.00	3.50
Corn meal	5.00	4.00	3.00	2.00
Barley	4.00	4.00	2.00	1.00
Soy lecithin	4.00	4.00	4.00	4.00
Fish gelatine	3.00	3.00	3.00	3.50
Potato concentrate	3.00	3.00	3.00	3.00
Fish oil	3.00	3.00	3.00	3.00
Brewer’s yeast	3.00	3.00	3.00	3.00
Linseed meal	2.50	2.50	2.50	2.50
Soy protein concentrate	2.17	2.17	3.17	3.17
Pre-digested fishmeal	2.00	3.00	3.00	3.00
Pea protein concentrate	2.00	2.00	2.00	3.00
Wheat gluten	1.00	2.00	4.00	5.00
Vitamin and mineral premix 3	1.00	1.00	1.00	1.00
Antioxidant	0.20	0.20	0.20	0.20
Sodium propionate	0.10	0.10	0.10	0.10
Vitamin E	0.03	0.03	0.03	0,03
Total	100	100	100	100
Proximate composition (on DM basis)
Crude protein (%)	58.84	58.36	58.53	58.42
Crude lipid (%)	13.61	14.77	15.73	16.75
Fibre (%)	0.99	3.04	4.91	6.92
Starch (%)	11.01	9.72	7.35	5.25
Ash (%)	8.07	7.09	6.15	5.13
Gross energy (MJ/kg)	20.91	20.91	20.77	20.62
Amino Acids (g/kg)
Arginine	3.74	7.32	10.7	14.34
Histidine	1.05	4.22	7.23	10.41
Isoleucine	2.03	5.85	9.47	13.30
Leucine	4.11	10.18	15.95	22.04
Lysine	3.88	8.75	13.32	18.22
Threonine	2.10	5.54	8.79	12.24
Tryptophan	0.59	2.10	3.53	5.03
Valine	2.18	8.23	13.96	20.02
Methionine + Cysteine	2.12	4.06	5.90	7.82
Phenylalanine + Tyrosine	4.42	5.26	6.12	6.97
Minerals and vitamins
Total p (%)	1.27	1.08	0.90	0.72
Ca (%)	1.82	1.53	1.26	0.96
Na (%)	0.76	0.50	0.26	0.01
K (g/kg)	5.90	3.89	2.01	-
Vit C (mg/kg)	1020.40	1020.41	1020.41	1020.41
Vit E (mg/kg)	255.10	255.10	255.10	255.10
Vit D (IU/kg)	2638.50	2456.12	2284.44	2102.04
Fatty Acids
DHA/EPA	1.23	1.18	1.09	0.93
DHA + EPA (%)	1.64	1.30	0.97	0.63

¹ Fishmeal (moisture: 6.8%; crude protein: 71.6%; crude lipid 8.2%; fiber: 0%; ash: 12.6%; gross energy: 20.4 MJ/kg). ² Black soldier fly larvae meal (moisture: 5%; crude protein: 57%; crude lipid: 14%; fiber: 12%; ash: 7%; gross energy: 18.9 MJ/kg). ³ Vitamin and mineral premix (kg of product): vitamin A = 1,200,000 IU; vitamin D3 = 200,000 IU; vitamin E = 12,000 mg; vitamin K3 = 2400 mg; vitamin B1 = 4800 mg; vitamin B2 = 4800 mg; vitamin B6 = 4000 mg; vitamin B12 = 4800 mg; folic acid = 1200 mg; calcium pantothenate = 12,000 mg; biotin = 48 mg; nicotinic acid = 24,000 mg; Mn = 4000 mg; Zn = 6000 mg; I = 20 mg; Co = 2 mg; Cu = 4 mg; and Se = 20 mg.

. The experimental diets were isoenergetic, isonitrogenous, and formulated to meet the specific nutritional requirements for Cyprinidae [30]. The feeds were supplied ad libitum, according to the “five minutes rule” described by Lawrence [31].

2.3. Data Collection

On days 0, 7, 14, 28 and 42, each fish was anesthetized by using a water solution containing 0.160 mg/mL of tricaine–MS-222 (Sigma^®, St. Louis, MO, USA). Afterward, the fishes were individually photographed using a digital camera, and body weight (BW) was recorded. The digital photographs of the fish were then used to identify each fish as described in previous studies [4,32]. Hence, BW gain (BWg) was calculated as the difference between the BW related to the two following time points and as the difference between the final and initial BW (BWf and BWi, respectively). The feed intake (FI) of each replicate was also recorded, and feed conversion rates (FCR) were calculated as FI/BWg. Furthermore, fish mortality was recorded on a daily basis. Feed palatability, fish behavior, and overall fish welfare were continuously evaluated to detect possible signs of disease or distress.

2.4. Tissue Sampling and Intestinal Morphometry Measurement

At the end of the experimental period (day 42), 16 fish per treatment (4 per replicate) were sacrificed by overdose of anesthesia (0.25 mg/mL, MS-222, Sigma^®, St. Louis, MO, USA). The fish were then dissected, and the entire digestive tract was removed, placed on acetylated support, fixed in buffered formalin solution at 10% (pH 7.4), and paraffin embedding was accomplished to obtain longitudinal sections of the intestine. Samples included in paraffin were then sliced into 5 µm sections and stained with hematoxylin and eosin to measure villus length. This parameter was obtained manually by tracing a segment along the longitudinal axis of each villus from its apex (i.e., the epithelial surface of the mucosa) to the tunica serosa, as already described in a previous study [4], where two independent measures were taken, named respectively upper and lower according to length (higher the upper compared to the lower). Sections of each intestine were also stained with Alcian blue (pH 2.5) and PAS (Periodic Acid-Schiff) to assess density (number/mm² of the mucosal surface) and area (µm²) of goblet cells. Sections were examined using a light microscope (Nikon, Eclipse 80i, Calenzano, Italy) connected to a PC via a Nikon digital system (Digital Sight DSU1). Representative images were acquired using NIS-Elements software to perform morphometrical analysis.

2.5. RNA Extraction from Tissues

The mRNA expression analysis was performed on 16 fish intestines per treatment. The whole intestines removed were wet weighted, immediately stored in a suitable volume of RNALater (Life Technologies) and frozen at −80 °C until RNA extraction. The intestine samples from fish were processed for RNA extraction by using the RNeasy^® Plus mini kit (Qiagen, MI, Italy) protocol, according to the manufacturer’s instructions, implemented with the on-column PureLink DNase (Qiagen) treatment to avoid genomic DNA contamination. Briefly, after the removal of RNALater excess, fat deposits and luminal debris, tissues were lysed with 400 µL Trizol reagent/sample until complete homogenization. At the end of the extraction protocol, RNA aliquots were kept stored in RNase-free conditions at −80 °C until use. RNA concentrations were calculated by spectrophotometry, and the λ260/λ280 ratios were calculated to evaluate possible protein contamination. All the RNA extractions were checked, qualitatively and quantitatively, by loading RNA samples on agarose gels.

2.6. Primer Design

Nucleotide sequences of reference mRNAs from the investigated genes were collected at the GenBank database (https://www.ncbi.nlm.nih.gov/, accessed on 6 June 2022). Gene-specific nucleotide tracts were selected as primer pairs for real-time PCR (qPCR) assays to perform expression analysis of the Zebrafish pept1, gata4, and nfkb1b genes; primer pairs were also designed for the Zebrafish ribosomal 28S rRNA chosen as housekeeping gene for normalization. By mRNA-to-genomic sequence alignment, the gene-specific oligonucleotide pairs of forward and reverse primers were invariably designed on different exons (intron spanning) to avoid amplification of genomic DNA. To simulate assay features of primer effectiveness (GC content, end stability, self/cross-dimer formation, and melting temperature) during amplification, the program AmplifX version 1.7.0 was used (https://inp.univ-amu.fr/en/amplifx-manage-test-and-design-your-primers-for-pcr, accessed on 6 June 2022). All primer pairs were tested for efficiency and specificity by RT-PCR before qPCR analysis. Amplification products corresponding to primers to be adopted in the subsequent qPCR analysis were sequenced and identified by alignment with reference sequences. Sequences and details of the pept1, gata4, nfkb1b and 28S-specific primers used for PCR assays are reported in

Table 2. Details of the primer sequences adopted for qPCR assays. Forward and reverse nucleotide sequences with melting temperatures (Tm) are reported, along with the reference Zebrafish genes and mRNA sequences (RefSeq mRNA) used for primer design. The PCR product size of the generated amplicons is also reported.

zf Genes	RefSeq mRNA	Sense Primer 5′–3′ (Tm)	Antisense Primer 5′–3′ (Tm)	PCR Size (bp)
slc15a1/pept1	NM_198064.1	TGTGACCATCTCTGCTGGAG (56 °C)	CCGCGTGCACATTATCAGAC (56 °C)	206
gata4	NM_131236.2	TCAAACCACAGAGACGACT (52 °C)	GTTGCAGACTGGCTCTCCTT (56 °C)	116
nfkb1b	XM_021481269.1	CACAGACAGTTTGCCATCGT (55 °C)	ATCTGTGGATGGTAGGTGAA (52 °C)	143
28S rRNA	EF417169.1	GGTCTAAGTCCTTCTGATGG (55 °C)	GGCTGCATTCCCAAACAAC (55 °C)	112

.

2.7. Reverse Transcription and qPCR

For each total RNA extraction, two reverse transcriptions were performed on 250 ng RNA each, using the Bio-Rad iScriptTM Select cDNA Synthesis kit (Bio-Rad, Segrate, Italy) according to the manufacturer’s instructions and in the presence of random primers. Primer efficiency in qPCR protocols for the expression analyses of pept1, gata4, nfkb1b and 28S were calculated according to the amplification efficiency parameters for genes of interest and internal controls proposed by Schmittgen and Livak [33]. Briefly, tenfold serial dilutions of reverse-transcribed cDNA (from control sample RNA) were used for qPCR in the presence of primers for each of the three investigated genes and the housekeeping control. Threshold Cycle (CT) output values (y-axis) were plotted vs. log cDNA dilution (x-axis) to determine the slope of the line. PCR efficiency was then calculated by the equation m = −(1/logE), where m is the slope of the line and E is the efficiency. Primer concentrations were optimized to obtain efficiencies fitting within 10% of each other. qPCR was performed using the IQ SYBR GREEN SUPERMIX protocol (Bio-Rad) with a CFX96 Touch^TM Real-Time PCR Detection System (Bio-Rad). In the qPCR analysis, gene expression relative quantification was calculated by analyzing the output threshold values (CT) by the comparative CT method (also referred to as the 2-ΔCT or 2-ΔΔCT method), as previously described (Livak and Schmittgen, 2001; Schmittgen and Livak, 2008), and qPCR data shown were the 2^−ΔCT values, which are considered to be proportional to the amount of detected target mRNA. ΔCT values (ΔCT = target gene CT—housekeeping gene CT) were obtained from 2 different rounds of qPCR (starting from 2 different retrotranscribed cDNA templates per biological replicate) for each target mRNA and the 28S internal (housekeeping) control.

2.8. Statistical Analysis

Survival rate data were obtained as the sum of the surviving fishes in all replicates for each treatment. Survival rate data of all treatments were simultaneously analyzed by Log-rank test using the product-limit (Kaplan-Meier) method for factors of right-censored data. Data related to BW, BWg, FI, FCR, and intestinal morphometries were tested for normal distribution using the Shapiro-Wilk test. Since data were normally distributed, mean values were tested by one-way ANOVA followed by Tukey’s HSD test. Since goblet cell density and area were not normally distributed, differences between treatments were detected using Wilcoxon non-parametric test. Differences were considered significant for p-value < 0.05. Data from qPCR assays were analyzed using one-way ANOVA, standard deviation, and standard error of the means. According to Schmittgen and Livak [33], statistical analyses were performed after the 2^−ΔCT transformation.

3. Results

No signs of abnormal behavior or distress were observed throughout the experimental period in fish on any treatment. Also, the feeds supplied were regularly consumed, and no feeding behavior differences were observed among treatments.

3.1. Survival Rate and Growth Performances

Fish survival rate was 95%, 98.5%, 100% and 93.6% for groups IM0, IM17, IM33 and IM50, respectively, and no differences were observed (p > 0.05). No differences were also observed for FI and FCR on days 7, 14, 28 and 42 of the experimental periods (p > 0.05), as well as for the cumulative FI. On the contrary, the cumulative FCR was significantly higher (p < 0.05;

Table 3. Growth performances of the fish fed with different Hermetia illucens meal (IM) dietary inclusion and FM replacement rates.

Parameters	IM0	IM17	IM33	IM50	SEM	p-Value
Feed intake (mg)
Day 0–7	21.0	23.0	22.0	23.0	0.001	0.916
Day 8–14	47.0	47.0	38.0	43.0	0.0028	0.619
Day 15–28	110.0	131.0	123.0	129.0	0.0063	0.664
Day 29–42	152.0	120.0	119.0	121.0	0.0057	0.170
Cumulative FI (day 0–42)	331.0	321.0	301.0	318.0	0.0066	0.490
Feed conversion rate
Day 0–7	1.0	0.9	0.9	1.0	0.071	0.883
Day 8–14	2.2	2.0	1.7	1.8	0.141	0.592
Day 15–28	1.8	1.8	1.8	2.0	0.085	0.943
Day 29–42	2.9	2.2	2.4	1.9	0.127	0.129
Cumulative FCR	2.1 a	1.8 b	1.8 b	1.8 b	0.032	0.013 *
Body weight gain (mg)
Day 0–7	21.5 b	25.9 a	26.2 ab	25.4 a	0.508	0.004 *
Day 8–14	21.2 b	23.7 ab	22.2 ab	24.9 a	0.454	0.029 *
Day 15–28	61.1 b	71.2 a	67.7 ab	67.7 ab	1.277	0.045 *
Day 29–42	52.6 b	55.6 ab	51.7 b	61.6 a	1.085	0.007 *
Cumulative BWg	156.5 b	176.8 a	168.0 ab	180.1 a	2.708	0.012 *
Body weight (mg)
Day 0	43.4	45.1	42.8	43.1	1.30	0.922
Day 7	65.0	71.0	68.7	68.6	1.61	0.626
Day 14	86.3	94.6	90.8	95.8	1.88	0.282
Day 28	147.6	165.8	158.5	164.3	3.06	0.144
Day 42	200.2	221.2	210.3	226.3	3.74	0.068

IM0-control: 0% IM; IM17: 17% IM; IM33: 33% IM; IM50: 50% IM. FI: feed intake, FCR; feed conversion rate, BWg: body weight gain and BW: body weight. SEM, standard error of the mean. Different lowercase letters indicate statistical differences between diets per p < 0.05 (see asterisk).

) in the control group (2.1) than in all the experimental diets (1.8). Cumulative BWg was significantly greater (p < 0.05) for treatments IM17 (176.8 mg) and IM50 (180.1 mg) in comparison to the control treatment (156.5 mg). No differences were observed between IM33 and the other treatments (Table 3).

3.2. Intestinal Morphometry

Villus length did not show any difference among treatments, neither in general nor in relation to the upper and lower side of the intestinal lumen (

Table 4. Villus length (µm) of the fish fed with different Hermetia illucens meal (IM) dietary inclusion and FM replacement rates. For each treatment, the villus length is also reported for the upper and lower surface of the intestinal lumen.

Parameter	IM0	IM17	IM33	IM50	SEM	p-Value
Villus length	165	160.5	171.0	161.0	0.51	0.813
Upper	170.3	173.5	178.5	171.5	0.51	0.737
Lower	164.4	145.8	161.7	152.9

IM0-control: 0% IM, IM17:17% IM, IM33: 33% IM and IM50: 50% IM. SEM, standard error of the mean. Different lowercase letters indicate statistical differences between diets per p < 0.05.

and Figure 1A). On the contrary, when the villi length was considered in relation to the upper and lower lumen surface and not considering the dietary treatment (

Table 5. Villus length (µm) was reported for the upper and lower surface of the intestinal lumen, independently from the dietary treatment.

Parameter	Upper	Lower	SEM	p-Value
Villus length	173.7 a	156.5 b	0.51	0.016

SEM, standard error of the mean. Different lowercase letters indicate statistical differences between diets per p < 0.05.

), the villi of the upper intestinal lumen were longer (173.7) than those on the lower side (156.5; p < 0.05).

No differences among treatments were observed for the density and area of Alcian blue-positive goblet cells (

Table 6. Intestinal morphometry in fish fed with different Hermetia illucens meal (IM) dietary inclusion and fishmeal (FM) replacement rates.

Parameter	IM0	IM17	IM33	IM50	SEM	p-Value
GC-AB density	486.2 SD 351.74 (n = 10)	711.5 SD 365.5 (n = 5)	333.4 SD 99 (n = 3)	499.5 SD 361.14 (n = 10)	3.61	0.307
GC-AB area	54.2 SD 16.84 (n = 10)	70.1 SD 19.5 (n = 5)	62.5 SD 0.84 (n = 3)	51.8 SD 9.51 (n = 10)	4.64	0.200
GC-PAS density	598 SD 280.86 (n = 9)	514.4 SD 136.4 (n = 4)	236.5 SD 147.94 (n = 4)	603.9 SD 291.87 (n = 9)	6.25	0.100
GC-PAS area	41.3 SD 8.51 b (n = 9)	53.6 SD 16.8 ab (n = 4)	63.3 SD 6.19 a (n = 4)	36.8 SD 10.58 b (n = 9)	11.29	0.010

IM0-control: 0% IM, IM17:17% IM, IM33: 33% IM and IM50: 50% IM. Number (n), density (number/mm² of the mucosal surface) and area (µm²) of Alcian blue positive (GC-AB) and PAS-positive (GB-PAS) goblet cell. SEM, standard error of the mean. Different lowercase letters indicate statistical differences between diets per p < 0.05.

and Figure 1B) and for the density of PAS-positive goblet cells. On the contrary, the area of PAS-positive goblet cells (Table 6 and Figure 1C) was significantly greater (p < 0.05) in the IM33 diet group (63.3 µm²) than in the control treatment (41.3 µm²).

3.3. Pept1, gata4 and nfkb1b mRNAs Expression in Intestine Tissues

In relation to the mRNA expression of pept1, gata4 and nfkb1b genes at the intestinal level, no differences were observed among the four dietary treatments (Figure 2). Nevertheless, the expression of gata4 mRNA revealed a positive dose-dependent trend of mRNA up-regulation (up to +2.2-fold change with respect to the control; fold change = 1).

4. Discussion

The lack of feeding behavioral differences among treatments suggests that all feeds had acceptable palatability, independently of the inclusion level of the Hermetia illucens meal in the feed. Moreover, in the present study, the inclusion level of the Hermetia illucens meal was higher than those used in other studies [4,34], and the meal used was full fat rather than defatted [34]; this aspect is quite relevant since fat might strongly impact the quantitative production of raw materials, the characteristics of the pellets, as well as the overall feed palatability.

The few mortality events observed in the present study were mainly recorded in the early stages of the experimental period when fish were still in a fast-growing phase and not yet in the adult stage. After that, mortality events were mainly associated with the procedures related to the periodic measurements (handling, anesthesia, etc.) and thus to the handling stress (i.e., capture, anesthesia, photographic detection, and weighing) rather than dietary treatments. Overall, the results suggested that the dietary inclusion of Hermetia illucens meal up to 50% (100% replacement of fishmeal) did not negatively affect fish survival rate and feed intake. On the contrary, FCR was significantly more favorable when Hermetia illucens meal was included even at the higher level, and totally replacing fishmeal. Similar results were observed in a previous study when lower inclusion rates (but similar fishmeal replacement rates) were used [4]. The only difference of interest in this work is related to the FCR values. When the dietary inclusion rate of Hermetia illucens meal was high, as in the present study (from 17% to 50% and from 33% to 100% fishmeal replacement rate), they decreased from 2 to 1.8. However, the FCR values observed were consistent with those reported in Zebrafish by Yossa et al. [35]. Also, Stejskal et al. [34] reported similar FCR values (1,81) in pikeperch (Sander lucioperca) when 36% of Hermetia illucens meal (defatted) was included in the diet; on the contrary, the diets characterized by a lower level of Hermetia illucens meal showed much more favorable FCR values (1,28 and 1,26 with 9 and 18% Hermetia illucens meal, respectively). This means that while Stejskal et al. [34] in pikeperch observed a negative correlation between Hermetia illucens meal (defatted) inclusion level, the results of the present study (carried out on zebrafish) suggest a positive effect of the inclusion of the Hermetia illucens meal (full fat) on the FCR.

In the lack of specific studies, this result can be hypothetically explained based on the dietary differences and notably with the slightly higher crude lipid content and lower fiber, starch and ash content in the diet containing HI meal. Similarly, another hypothesis might be related to the dietary amino acids profile, notably for Lysine and Methionine and Cysteine, among others (Table 1).

Partial or total fishmeal replacement with insect meal has demonstrated no negative effects on feed utilization and mortality in several other species of interest for the aquaculture sector [36,37]. Kroeckel et al. [17] reported that the dietary inclusion of Hermetia illucens meal from 17 to 33% for replacing fishmeal showed no negative effect on survival, growth, and feed utilization in turbot. Other studies [11,19] reported similar results on other species, such as Atlantic salmon [18], Gilthead seabream (Sparus aurata), European seabass (Dicentrarchus labrax) and Japanese seabass (Lateolabrax japonicus). Furthermore, several studies reported that fishmeal replacement with Hermetia illucens meal up to 50% resulted in similar growth performances and feed utilization in Rainbow trout (Oncorhynchus mykiss) [16,24,36,38,39], Jian carp (Cyprinus carpio var. Jian) [22] and Nile tilapia (Oreochromis niloticus) [40]. Moreover, comparable yield data with no significant alterations (i.e., off-flavors) were reported by several authors in trout and salmon meat [16,18].

Similar results [41,42] were also obtained by replacing fishmeal with Tenebrio molitor meal in some fish species, such as Rainbow trout (Oncorhynchus mykiss) and Blackspot seabream (Pagellus bogaraveo). In Blackspot seabream [42], no significant decrease was recorded for FI, FCR, BWg and fillet characteristics (i.e., texture properties and water holding capacity). In Rainbow trout (Oncorhynchus mykiss), the feeding rate was lower in the group fed a diet including 50% Tenebrio molitor meal and 25% fishmeal compared to the group fed a fishmeal-based diet (control; 75% fishmeal). This lower FI did not negatively affect the BWg but rather resulted in better FCR and SGR, as well as better PER (protein efficiency ratio). Also, the survival rate followed the same trend, being higher for the groups fed Tenebrio molitor meal [41].

Notably, insect meal protein inclusion in fish diets might also either reduce or improve fish growth performance and feed utilization, essentially depending on its inclusion level and the fish species. In turbot, according to Kroeckel et al. [17], Hermetia illucens meal inclusion level higher than 33% (from 49 to 76%) demonstrated a decrease of BWg, FI and FCR. Similarly, Xiao et al. [12] observed a BWg reduction in Yellow catfish when the fishmeal replacement rate was higher than 65%. Moreover, low FCR and growth performance [15] were also observed in other fish species fed a high level of Hermetia illucens meal, such as Channel catfish (Ictalurus punctatus) and Blue tilapia (Oreochromis aureus), but some doubts may arise in relation to the characteristics of the meal production process.

Concerning the aspects related to Zebrafish growth performances, the results observed in the present study differed from those reported in a previous one [4]. The results also differed from those described by Kroeckel et al. [17] on turbot. Cumulative BWg was significantly higher in treatments IM17 and IM50, in comparison to control (IM0), and FCR resulted higher in the control treatment than in the experimental treatment. In addition, fish final BW did not result in significant differences between the control and experimental treatments. However, the highest final BW (not significantly) was recorded in treatments IM17 and IM50. Therefore, these results suggest that in Zebrafish level of Hermetia illucens, meal inclusion higher than 33% lead to better growth performance.

In accordance with other studies carried out on Zebrafish by Zarantoniello et al. [43,44], Vargas et al. [45], and Fronte et al. [4], intestinal morphometry did not show any histological alteration and significant differences in villus length of zebrafish. This result might confirm the hypothesis that this ingredient is devoid of allergens and other substances that are detrimental to enterocytes, as suggested by Li et al. [22]. The absence of intestinal histological alteration was also previously reported in several fish species fed a diet partially or totally composed of Hermetia illucens meal, such as Atlantic salmon, Rainbow trout (Oncorhynchus mykiss) and Japanese seabass [18,23,24,46]. In Jian carp, pathological intestine damage (i.e., tissue disruption) was observed when the fishmeal diet was replaced with insect meal for more than 75% [22]. Black soldier fly meal inclusion higher than 50% (75 and 100%) also showed hepatic steatosis, microbiota modification, higher lipid content, fatty acid modification and higher expression of immune response markers in zebrafish [44]. Therefore, as previously suggested by Li et al. [22], the results obtained in this study seem to confirm that a 50% inclusion level of Black soldier fly meal in the diet for zebrafish represented the best compromise between ingredient sustainability and proper fish growth and welfare.

Data available in the scientific literature on insect meal inclusion effect on intestine histology are still yet variable and not exhaustive. In Rainbow trout (Oncorhynchus mykiss), a significant reduction of villus length in the anterior intestine was observed when fed a diet containing partially defatted Black soldier fly larvae meal at 26.4% inclusion level (100% fishmeal replacement). Morphological changes in the anterior intestine were also detected in Jian carp-fed defatted Black soldier fly larvae meal at 8% inclusion [22].

Noteworthy is that, as already reported in a previous study [4], villus length in the upper intestinal lumen surface was significantly higher than those in the lower one. However, no differences were recorded in villus length in the upper and lower intestinal lumen surface among the considered groups, control included (no Black soldier fly meal). This suggests that the differences in villus length between the upper and the lower intestinal lumen surface of zebrafish are not diet related. Further investigation to assess whether these differences are due to an anatomical characteristic of zebrafish or to other factors (i.e., age or rearing system) is needed.

Concerning goblet cells, no differences were recorded among treatments for density and area of Alcian blue-positive goblet cells as well as for density of PAS-positive goblet cells. Conversely, the area of PAS-positive goblet cells resulted significantly higher in the IM33 diet group than in the control group and IM50. The intestinal barrier efficiency depends on host-microbial balance as well as mucin production [47,48,49]. Other studies showed that diet composition could strongly affect mucins production by goblet cells in European seabass (Dicentrarchus labrax), Common carp (Cyprinus carpio) and Gilthead seabream (Sparus aurata) [21,50,51,52]. However, although it seems that diet plays a key role in goblet cell size and density, based on the available knowledge, it is not possible to formulate a specific hypothesis for explaining the mechanism that underlies this finding. Hence, further investigations on the biological response of the digestive tract to different diets are necessary.

Because of its relevance in animal nutrition, Slc15a1/pept1 is one of the best-studied solute carriers amongst those characterized in teleost fish [53,54,55]. It is thus adopted as a marker of functional and regulatory expression of the correct setup of the absorptive potential of the intestinal epithelium in all vertebrates, teleost included. Due to its importance at the intersection of the alimentary functions of the gut, its response under physiological and dietary solicitations, and its possible involvement in examples of total body plasticity, such as growth and compensatory growth, pept1 represents a highly specific sensor [56]. On these bases, the results of the present study suggest that the considered diets were comparable in relation to dietary effectiveness and, particularly, to digestion and absorption of the protein components.

Besides pept1, data on gata4 transcriptional expression are coherent according to the equivalence of standard and insect meal treatments. Gata4 is a key regulator of epithelial cell differentiation and epithelial-mesenchymal interactions in the gut epithelium of zebrafish [57]. Gata4 is not significantly regulated by fishmeal replacement with insect meal in the Zebrafish gut, thus indicating no specific impact on the epithelium. Nevertheless, the results hint at an up-regulating trend of gata4 expression consequent to an increase of insect meal inclusion; interestingly [58], gata4 is a crucial “housekeeping” of the barrier function of the gut epithelium (i.e., it controls mucosal integrity) in vertebrates. In this view, the insect meal dietary assumption might be furtherly investigated to assess the ability to improve the barrier performance of intestinal epithelium after long-term feeding, which in turn might improve the healthy metabolism of fish.

In this regard, an added value comes from the assessment of the invariant expression of nfkb1b mRNA, whose protein product is a component of the NF-kB complex. In fact, NF-kB is a key node of innate immunity and inflammatory pathways in the fish gut, zebrafish included [59,60]. Overall, the results of the gene expression suggest a safe use of Black soldier fly meal and possible ameliorating effects at the intestine epithelium level and consequently on the whole fish’s health and growth performances.

5. Conclusions

The result of this investigation confirms the possibility of partially or totally replacing fishmeal with Hermetia illucens meal in Zebrafish diets, also when it is included up to 50% (and 100% fishmeal replacement). In fact, in this case, no negative effects on survival, feed intake, feed conversion rate, or growth performance were observed. In addition, any detectable histological alterations at intestinal level were observed. Similarly, the absence of significant alterations of the transcriptional expression of genes marking intestinal epithelium functionality, such as pept1, nfkb1b, and gata4, confirm and support this opportunity.

Since zebrafish is a key experimental model in biology (e.g., from biomedicine to aquaculture), the study provides additional knowledge on the potential use of Black soldier fly meal, also in fish of interest for the aquaculture sector. In turn, it is necessary to consider its high market price, that nowadays is not yet economically competitive with fishmeal, but, in the perspective of the increasing demand for seafood, the fishmeal market shortage, the increasing insect meal production, and the needs of improving aquaculture "sustainability", insect meal possess all the required characteristics to become a convenient alternative to fishmeal for aquafeed production [61]. However, further investigations on nutritional aspects related to other fish species, as well as the socio-economic impact of the use of this alternative protein source, are suggested.