An enteric ultrastructural surface atlas of the model insect Manducasexta

Summary The tobacco hornworm is a laboratory model that is particularly suitable for analyzing gut inflammation, but a physiological reference standard is currently unavailable. Here, we present a surface atlas of the healthy hornworm gut generated by scanning electron microscopy and nano-computed tomography. This comprehensive overview of the gut surface reveals morphological differences between the anterior, middle, and posterior midgut, allowing the screening of aberrant gut phenotypes while accommodating normal physiological variations. We estimated a total resorptive midgut surface of 0.42 m2 for L5d6 larvae, revealing its remarkable size. Our data will support allometric scaling and dose conversion from Manduca sexta to mammals in preclinical research, embracing the 3R principles. We also observed non-uniform gut colonization by enterococci, characterized by dense biofilms in the pyloric cone and downstream of the pylorus associated with pore and spine structures in the hindgut intima, indicating a putative immunosurveillance function in the lepidopteran hindgut.


INTRODUCTION
Preclinical research has long been reliant on small mammals such as mice, but heightened ethical considerations have prompted a shift toward the 3R principles (replacement, reduction, and refinement) in regulations governing animal experiments and research funding. 1,2In this evolving context, insect larvae have emerged as promising alternative in vivo animal model systems, not only because certain physiological aspects are remarkably similar to humans but also because they can be bred and reared in large numbers at very low costs. 3,4he tobacco hornworm (Manduca sexta) is an insect model organism with many advantages for preclinical research, including a fully mapped genome, 5,6 methylome, 7 and the availability of specific monoclonal antibodies. 8The large size of M. sexta larvae, regularly exceeding 10 g, is also ideal for experiments in biochemistry, 9 developmental biology, 10 immunology, 9,11,12 epigenetics, 7 morphology, 13 neurobiology, 14 and gut physiology. 15There is a high degree of evolutionary conservation between the M. sexta and human gut, including notable similarities in enteric epithelial structures and innate immunity. 13,16ased on these advantages, we previously established M. sexta as a non-vertebrate model of gut inflammation that adheres to the 3R principles. 16We have used medical imaging modalities such as computed tomography (CT), 13 magnetic resonance imaging (MRI), 17 and positron emission tomography (PET), 18 exploiting the large size of M. sexta larvae for the image-guided high-throughput screening of aberrant gut phenotypes, including screening for new effectors and inhibitors of gut inflammation, pesticides, antibiotics, and host-pathogen interactions, 16 as well as the identification of new contrast agents for medical imaging. 191][22][23][24][25] However, there are anatomical and physiological differences between parts of the gut of M. sexta that are not pathological, so it is important to have a standard reference in healthy insects that can be used for comparative purposes.][28][29] In particular, this study complements our previously published quantitative three-dimensional micro-tomographic gut atlas of M. sexta. 13e also considered the transition from murine to insect models by calculating the total resorptive midgut surface area, allowing allometric scaling for oral dose conversion, using our previously established mCT method and a clinical contrast agent.This was sufficient for the overall volume calculations and revealed the most prominent gut folds, but to quantify the deep villus-like gut folds, we combined an advanced nano-and micro-computed tomographic (CT) approach with SEM.Previously, we revealed a striking similarity in gut volume between

Foregut
The intima of the foregut is heavily folded and smooth (Figures S1, S2, and Video S1).Previously, we showed that the intima of the foregut is highly extendable and can form a crop. 13The foregut opens into the midgut, marked by the presence of the stomodeal valve, which extends into the midgut lumen. 13The surface of the stomodeal valve is similar to that of the foregut (Figure S3).Previously, we reported the volume of the L5d6 foregut as 0.03 ml, with a mean area of 68.57mm 2 . 13

Midgut
Based on its general morphology, the M. sexta midgut can be subdivided into the anterior, middle, and posterior midgut. 13,29Previously, we showed that six rudimentary caeca are located at the most anterior part of the anterior midgut. 13In this region, we observed elongated cylindrical structures of unknown function in an annular arrangement, which do not occur elsewhere in the intestine (Figure S4).In our previous paper, we documented a midgut volume of 1.4 ml. 13

Midgut epithelium and peritrophic matrix
The midgut is the primary site of nutrient resorption.Microvilli are membrane protrusions that increase the surface area for the absorption or secretion in most cells in the midgut.In contrast to the microvilli in the middle and posterior regions, the microvilli in the anterior midgut show extensive thickening and appear as club-shaped microvilli (Figure 1). 29The cell boundaries of the enterocytes are visible, and the peritrophic matrix has a felt-like structure, with pronounced nodules (Figure 1D; Video S2).The shape of the microvilli changes gradually in an anterior-to-posterior gradient (Figures 1, 2, and 3).The microvilli in the middle anterior midgut remain condensed and adopt their familiar shape only in the rear anterior midgut (Figure 3; Video S3).The two most common cell types in the gut epithelium, the columnar cells and goblet cells, are visible in cross-sections of the gut epithelium (Figure 2; Video S3).When the peritrophic matrix is completely intact, it covers the entire intestinal epithelium like a cloth.Filaments are visible in the topological canyons of the underlying epithelium (Figure 4; Video S4).The entire peritrophic matrix on the endoperitrophic surface is lightly colonized with bacteria (Figure 4; Video S4).When the peritrophic matrix is partially or completely detached, bacteria are no longer visible (Figures 5, S5, and Video S5), with very few exceptions (Figure 6C).The bacteria are present in the midgut and hindgut.They have an ovoid shape and appear to occur in pairs or chains of different lengths (Figures S7 and Figure 8E).According to our previous characterization of the M. sexta microbiome, these bacteria belong to the genus Enterococcus. 16,31The enterocytes of the posterior midgut gently protrude into the gut lumen.The peritrophic matrix can easily detach from the posterior midgut region and is not shown (Figure 6; Video S6).

Estimation of the total surface area of the midgut
We used mCT to determine the mean shrinkage-corrected surface area of the midgut (without deep, villus-like midgut folds and microvilli) resulting in a value of 0.0013 G 0.0004542 m 2 (n = 10) (Table 1, Figures 7 and S6).We then repeated the measurement, combining mCT and nanoCT, resulting in a value of 0.00401 G 0.001275 m 2 (n = 10).The shrinkage-corrected length of the midgut determined by mCT was 46.91 G 7.421 mm (n = 10).The microvilli density determined by SEM was (31.80 G 3.846)/mm 2 (n = 3).The mean microvillus diameter determined by SEM was 0.116 G 0.02242 mm (n = 3), and the mean microvillus length determined by SEM was 9.036 G 1.382 mm (n = 3).This equates to a mean microvillus surface area of 3.278 G 0.821 mm 2 (n = 3), a microvillus amplification factor (MAF) of 103.0 G 26.57 (n = 3), and a relative intestinal surface area (RISA) of 85.17 G 17.12 (n = 10) (Table 1, Figures 7 and S6).The estimated total surface area of the midgut was therefore 0.4242 G 0.1455 m 2 (n = 10) (Table 1) with mean surface area of 885 G 184.5 cm 2 per cm midgut length (n = 10) (Table 1).

Hindgut
The transition from the posterior midgut to the first part of the hindgut, the pyloric cone, is marked by an abrupt change in surface structure (Figure 8; Video S7).The cuticle-coated intima of the pyloric cone abruptly displaces the enterocytes mounted with microvilli.Unlike the peritrophic matrix, the pyloric cone is densely populated with enterococci.Next, the pylorus shows the characteristic armature of spiculated pads at the pyloric valve (Figure S8). 32,33The dense population of bacteria persists through the ileum and the cuticle surface features numerous spikes (Figures 9, S9, and Video S8).Densely spiculated fields alternate with spines arranged in lines (Figure S9).In general, the direction of the lines is not uniform, and in some cases, they form perpendicular arrays, suggesting a role in shredding the passing peritrophic matrix (Figure S9).At higher magnification, pores in the intima become evident (Figure 9; Video S8).The colon is heavily folded and devoid of spines and bacteria (Figure 10; Video S9), but the intima still features pores.The intima at the colon-rectum transition shows spherical imprints or sacculations (Figure S10).Also, significant amounts of debris and remains of the peritrophic matrix are present in this area.Finally, the rectum shows a smooth and lightly folded intima without spines or pores but with significant amounts of debris (Figures 11H-11J, S11, and Video S10).We previously determined the volume of the L5d6 hindgut (without the pyloric cone), reporting a value of 0.208 ml with a mean area of 348.2 mm 2 . 13

DISCUSSION
We have developed a comprehensive gut surface atlas of the model insect Manduca sexta, which documents the physiological standard of healthy animals as a reference for the qualitative phenotyping of gut-associated pathologies such as gut inflammation. 16The atlas also allowed us to estimate the resorptive surface area of the midgut, which will facilitate comprehensive allometric scaling for enteral dose conversion from insects to mammals.

Foregut
The foregut has the primary functions of pre-digestion, food storage (as a crop), supplying food to the midgut.The foregut also facilitates regurgitation, an essential defensive strategy against predators. 34,357][38][39][40][41] In line with its function, the cuticle of the M. sexta foregut has a very low permeability compared to the cuticle of the ileum and rectum. 42

Midgut
The midgut is the principal site for digestion and nutrient resorption. 13,29,34The primary cell types found in the M. sexta midgut epithelium (and that of other lepidopterans) are columnar cells and goblet cells.The columnar cells are responsible for the secretion of digestive enzymes and the absorption of nutrients, whereas goblet cells actively pump K + from the hemolymph into the gut lumen using a K + /2H + antiporter mechanism, thereby maintaining an alkaline environment in the midgut as is typical for most lepidopterans. 13,34,43olumnar cells and goblet cells undergo significant morphological changes along the anterior-posterior axis. 13,29,34In the anterior midgut, columnar cells have club-shaped microvilli on their apical surfaces, and cytoplasmic bridges connect them to form a complex meshwork.Regular microvilli are found only from the rear of the anterior midgut toward the middle midgut. 13,29,34In the anterior midgut, goblet cells feature a widened cavity located at the cell base with a long neck connecting it to the lumen.Microvilli lining this cavity contain mitochondria, and the nucleus is at the level of the widened cavity. 29In contrast, goblet cells in the posterior midgut feature a cavity restricted to the upper part of the cell above the nucleus, giving them a goblet shape.Microvilli in this region lack mitochondria. 29Both types of goblet cell have a valve of closely spaced microvilli at the cavity entrance. 44These differences are not graded but change abruptly between midgut regions in M. sexta. 13,29,34Such physiological variations should not be confused with the symptoms of gut inflammation and infection. 16Because goblet cell microvilli (projections of the goblet cell cave membrane) are not part of the resorptive surface, we did not include them in the quantification of the midgut surface area.Extensive epithelial erosion, a hallmark of conditions such as gut inflammation and infectious colitis in both mammals and insects, is easy to recognize by SEM. 16,45Changes in plasma membrane dynamics and enterocyte morphology are typical of these conditions, including enterocyte blebbing, 46,47 shedding 23,46 and purging 20 as well as the shedding or expulsion of microvilli. 20,21,47Previously, we and others documented these SEM-accessible phenotypes in M. sexta and observed enterocyte swelling, shedding and membrane extrusion 16,27 as well as blebbing and necrosis 16 augmented by the shortening and loss of microvilli. 16,26,27ata concerning the resorptive midgut surface are needed for allometric dose scaling from caterpillars to mice in preclinical studies.We therefore combined our SEM, mCT and nanoCT data to estimate the total resorptive midgut surface area of L5d6 M. sexta larvae, revealing an average value of 0.42 m 2 (Table 1) although one particularly large specimen had a resorptive midgut surface area of 0.71 m 2 (Figure 7F).The average midgut surface area value of M. sexta corresponds to approximately one-third of the surface area of the mouse intestine. 30This is impressive because the mouse intestine is more than 10 times longer.We also found that M. sexta lacks intestinal villi and has a lower density of microvilli compared to mammals. 48However, this is partially compensated by the presence of microvilli that are nine times longer than those in mice. 29Accordingly, the MAF of M. sexta is four times higher than that of mice.Instead of villi, deep villus-like midgut folds are present in M. sexta.This enhances and refines our previous studies, which showed comparable intestinal volumes in mice and late-instar M. sexta larvae. 130][51] Nevertheless, our data will help to improve the assessment of allometric scaling.The peritrophic membrane is another central feature of the midgut, 13,52 consisting of a mucinous intestinal lining reinforced with chitin. 3,53he M. sexta peritrophic matrix consists of 60% protein and 40% chitin. 54Insect intestinal mucin (IIM) and chitin-binding proteins connect the chitin fibrils and form a gel-like membranous structure. 53By convention, two different types of peritrophic matrices are recognized in insects. 52ype 1 matrices are secreted from extensive areas of the midgut epithelium and consist of multiple laminae, 52 whereas type 2 matrices are produced by specialized cells of the stomodeal and mesenteric epithelia and consist of only a single layer. 52By the same convention, M. sexta has a type 1 peritrophic matrix. 52,55,56Interestingly, IIMs that have been described thus far share many similarities with MUC2, one of the core components of mucus in the human intestine and a key player in the pathogenesis of ulcerative colitis. 53,57,58][60] We found that the endoperitrophic surface of the peritrophic matrix in M. sexta, especially the pyloric cone and ileum, are populated by enterococci.Previously, we characterized the M. sexta gut microbiome and found the community was dominated by two species of enterococci, both acting as protective symbionts. 16Given their location and morphology, 31 the bacteria detected on the peritrophic matrix appear to be the same species we previously characterized.Importantly, these bacteria were present on the endoperitrophic sheet but not in the ectoperitrophic space in M. sexta, agreeing with other studies claiming that the peritrophic matrix is an effective bacterial barrier. 61he main structural difference between insect and mammalian intestinal mucins is the presence of chitin in the peritrophic matrices of insects. 53Human intestinal mucus forms a gel-like, shapeless lining within the intestine, with a typical thickness of 50-200 mm. 58In contrast, the peritrophic matrix in M. sexta and other insects is considerably thinner, typically 1-12 mm. 13,55,56Therefore, reinforcing this thinner mucin layer with chitin fibrils may help to achieve mechanical strength and stability equivalent to that observed in mammals. 53

Hindgut
Toxic nitrogenous waste from the Malpighian tubules is secreted into the lumen of the pyloric cone. 13The predominant waste product in M. sexta is uric acid. 62Some enterococci, such as Enterococcus faecalis, can metabolize uric acid. 63,64It therefore seems plausible that these bacteria have densely populated the pyloric cone and the entire hindgut downstream to the pylorus.On closer inspection, we found pores in the areas of the hindgut colonized by bacteria.To our knowledge, such pores have not been described in other lepidopteran species and their function is unclear.The main site of water reabsorption in M. sexta is the water-permeable intima of the rectum, 65 but we did not observe any pores there, making a potential role in the augmentation of water reabsorption unlikely.Because the pore-lined region, apart from the colon, is densely colonized with bacteria, the pores may help the underlying epithelium to control the bacterial population and facilitate an immune response if necessary.This hypothesis is supported by the high DUOX expression level in the larval hindgut and the secretion of prophenoloxidase. 66,67We also observed spines in the hindgut intima organized in pads, lines, and fields, which may help to grind the incoming peritrophic matrix, exposing the bacteria from the endoperitrophic space to the hindgut immunosurveillance.

Summary
Our comprehensive SEM-based atlas provides a detailed overview of the ultrastructure of the gut surface in M. sexta.Given its role as a model of gut inflammation and host-microbe interactions, we have developed a reference standard of healthy M. sexta individuals, identifying distinct morphological changes across the anterior, middle and posterior midgut.This provides a valuable tool for the SEM-based screening of abnormal gut-related traits such as inflammation and infection-induced colitis.Additionally, we estimated the total resorptive midgut surface area in L5d6 M. sexta larvae.The surface area of 0.42 m 2 highlights the remarkable size of the gut despite the absence of villi, relying instead on deep villus-like folds housing exceptionally long microvilli.These data are necessary for allometric scaling and accurate dose conversion from M. sexta to mammals in preclinical studies, while supporting the 3R principles.We also revealed the unique colonization pattern of enterococci, forming dense biofilms in the pyloric cone and downstream of the ileum, accompanied by distinctive pore and spine structures within the hindgut cuticle.These findings suggest the pores and spines may be involved in the immunosurveillance of the underlying gut epithelium.

Limitations of the study
This study reports a reference surface atlas of M. sexta larvae in the fifth larval stage on development day 6 (L5d6).Therefore, the reader should draw conclusions for developmental stages not represented in this study with caution.Furthermore, the impact of the animals' sex on the intestinal ultrastructure could not be ascertained due to the uncertainty of sex determination in M. sexta larvae.In addition, SEM analysis  can be prone to shrinkage artifacts.Therefore, additional care is needed in the evaluation of surface details at high magnifications (at and beyond 10,0003).

STAR+METHODS
Detailed methods are provided in the online version of this paper and include the following:

Figure 1 .
Figure 1.Anterior midgut of M. sexta A dense peritrophic matrix covers the gut epithelium.Enterocytes show irregular club-shaped microvilli.The cell boundaries are visible.A micro-tomographic surface overview of the digestive system of M. sexta (A) shows the localization (b 0 ) of the SEM insets (B-D).The image in panel (B) is artificially colored to highlight the peritrophic matrix.

Figure 2 .
Figure 2. Anterior midgut of M. sexta (epithelial cross-section)The gut epithelium consists of columnar cells and goblet cells.The irregular club-shaped microvilli gradually approach their standard shape in the anterior midgut toward the middle midgut (Figure3).A micro-tomographic surface overview of the digestive system of M. sexta (A) shows the localization (b 0 ) of the SEM insets (B-E).The image in panel (B) is artificially colored to highlight the goblet cells.

FFigure 3 .
Figure 3. Anterior midgut of M. sexta The midgut folds are covered with the peritrophic matrix.The microvilli have reached their standard shape.A micro-tomographic surface overview of the digestive system of M. sexta (A) shows the localization (b 0 ) of the SEM insets (B-F).The image in panel (B) is artificially colored to highlight the peritrophic matrix.

Figure 4 .
Figure 4. Anterior midgut of M. sexta with an intact peritrophic matrix Note the bacteria colonizing the peritrophic matrix on the endoperitrophic surface.A micro-tomographic surface overview of the digestive system of M. sexta (A) shows the localization of the SEM inset (B and C).

Figure 5 .
Figure 5. Middle midgut of M. sexta The prominent midgut folds of the gut epithelium are covered by the peritrophic matrix.Microvilli appear under the peritrophic matrix.A micro-tomographic surface overview of the digestive system of M. sexta (A) shows the localization (b 0 ) of the SEM insets (B-D).The image in panel (C) is artificially colored to highlight the peritrophic matrix.

Figure 6 .
Figure 6.Posterior midgut of M. sexta The enterocytes gently protrude into the posterior gut lumen.The peritrophic matrix can easily detach from the posterior midgut region and is not shown.A few bacteria are directly exposed to the microvilli, as shown in panel (C).The image in panel (C) is artificially colored to highlight the bacteria on top of the microvilli.A micro-tomographic surface overview of the digestive system of M. sexta (A) shows the localization (b 0 ) of the SEM insets (B-D).

Figure 7 .
Figure 7. Quantification of the total surface area of the M. sexta midgut (A-C) Quantification of the mean length (A), diameter (B) and density (C) of microvilli by SEM.(D) Quantification of the midgut surface area by mCT.(E) Quantification of the surface area of deep villus-like midgut folds by nanoCT.(F) The midgut surface area is dependent on the specimen weight.The upper and lower dashed lines represent the 95% confidence interval.(G-I)Boxplots of the (G) relative intestinal surface area (RISA), (H) microvillus amplification factor (MAF), and (I) total surface area of the M. sexta midgut.Boxplots show the 25 th to 75 th percentiles, with whiskers extending to the minimum and maximum data values while including all data points.The center denotes the mean, and the center line signifies the median.For a detailed list of all findings, please refer to Table1.

Figure 8 .
Figure 8. Hindgut (pyloric cone) of M. sexta In sharp contrast to the midgut, the folded intima of the pyloric cone (hindgut) is densely covered with a bacterial biofilm.A micro-tomographic surface overview of the digestive of M. sexta (A) shows the localization (B 0 ) of the SEM insets (B-D).The images in panels (C-E) are artificially colored to highlight the bacteria.A larger version of the image in panel (E) is provided as Figure S7.

Figure 9 .
Figure 9. Hindgut (ileum) of M. sexta The folded intima is spiculated and densely populated with bacteria.At higher magnification, pores in the intima become evident.A micro-tomographic surface overview of the digestive system of M. sexta (A) shows the localization (B 0 ) of the SEM insets (B-F).The image in panel (F) is artificially colored to highlight the bacteria.

Figure 10 .
Figure 10.Hindgut (colon) of M. sexta The heavily folded intima of the colon lacks spines and bacteria.The intima has spherical imprints (sacculations) at higher magnification and pores are visible.A micro-tomographic surface overview of the digestive system of M. sexta (A) shows the localization (b 0 ) of the SEM insets (B-E).The image in panel (E) is artificially colored to highlight the pores.

Table 1 .
Calculated features of the midgut surface area, including the total surface area of the midgut Casteleyn, C., Rekecki, A., Van der Aa, A., Simoens, P., and Van den Broeck, W. (2010).Surface area assessment of the murine intestinal tract as a prerequisite for oral dose translation from mouse to man.Lab.Anim 44, 176-183.10.1258/la.2009.009112. a