Vector-borne Trypanosoma brucei parasites develop in artificial human skin and persist as skin tissue forms

Transmission of Trypanosoma brucei by tsetse flies involves the deposition of the cell cycle-arrested metacyclic life cycle stage into mammalian skin at the site of the fly’s bite. We introduce an advanced human skin equivalent and use tsetse flies to naturally infect the skin with trypanosomes. We detail the chronological order of the parasites’ development in the skin by single-cell RNA sequencing and find a rapid activation of metacyclic trypanosomes and differentiation to proliferative parasites. Here we show that after the establishment of a proliferative population, the parasites enter a reversible quiescent state characterized by slow replication and a strongly reduced metabolism. We term these quiescent trypanosomes skin tissue forms, a parasite population that may play an important role in maintaining the infection over long time periods and in asymptomatic infected individuals.


Supplementary Note: Improved mechanical properties reduce dermal contraction and weight loss in primary human skin equivalents
A major problem with collagen-based skin equivalents is their susceptibility to fibroblast-mediated contraction during culture.The poor mechanical stability of the dermal component 1 makes standardization difficult.In this study, we refined a skin equivalent 2 by improving the mechanical stability of the dermal component to reach a high level of standardization and reproducibility.A computer-assisted compression system consisting of a custom-made compression reactor (Supplementary Fig. 1a, b) and a linear motor (Supplementary Fig. 1c) was used to compress the dermal component by a factor of 7. The high-density dermal equivalents (hdDEs) were characterized by a diameter of 12 mm, a height of 1 mm, and a calculated collagen concentration of 46.9 mg/ml.The increased collagen concentration and reduced water content improved the mechanical properties of the hdDEs, as shown by rheology.Strain sweep measurements confirmed a 3-fold higher storage modulus (G′) in the linear viscoelastic region of cell-free hdDEs compared to cell-free, non-compressed low-density dermal equivalents (ldDEs; ccollagen = 6.7 mg/ml), indicating stronger viscoelastic strength of the cell-free hdDEs (Fig. 1b and Supplementary Fig. 1d).The viscoelastic strength was even more pronounced in hdDEs colonized with normal human dermal fibroblasts (NHDF) for 23 days.Furthermore, ldDEs showed an earlier flow point (G′ = G′′) than hdDEs, correlating with a lower structural resistance against shear forces (Supplementary Fig. 1d).Compared to human skin biopsies, G´ was 11-fold lower in cell-free ldDEs, while only 3-fold lower in hdDEs (Fig. 1b and Supplementary Fig. 1d).In contrast to hdDEs, skin biopsies additionally included the epidermal skin layer, indicating an even smaller difference in mechanical stability between pure human dermis and hdDEs.Long-term comparison of hdDEs and ldDEs confirmed that hdDEs show significantly less contraction and weight loss over time, while having a similar number of proliferative NHDF (Supplementary Fig. 1e, and Supplementary Fig. 2a, b).While hdDEs lost only 6.7 % of weight and 15.7 % of area during 3 weeks of culture, ldDEs lost more than 82.7 % and 71.8 %, respectively.Scanning electron microscopy confirmed proper assembly of collagen fibers in both DEs (Supplementary Fig. 2c).However, in contrast to ldDEs, the collagen fibers in hdDEs showed a denser and more uniform distribution.

High-density skin equivalents recapitulate key anatomical, cellular, and functional aspects of native human skin
High-density dermal equivalents were colonized with normal human epidermal keratinocytes (NHEK) on day 3 and serial histological analysis showed that the morphology of high-density skin equivalents (hdSEs) on day 15 resembled that of native human skin, with a fully differentiated epidermis in which the 4 characteristic layers had formed (Fig. 1c, a-d).The epidermis was stable until day 23, and a prolonged culture was possible but ultimately led to a stratum corneum with aspects of hyperkeratosis.Thus, in a time window of minimally 9 days, the tissue architecture of the hdSE resembled native human skin.In addition, a comparative analysis showed that physiological skin markers were expressed in the hdSE at day 23 (Supplementary Fig. 2d).The presence of epidermal differentiation markers such as KRT5 and KRT14 in the basal layer, KRT10 in the suprabasal layers, and loricrin, filaggrin, and involucrin in the stratum granulosum and stratum corneum verified an anatomically correct differentiation of the epidermis.The presence of collagen IV demonstrated the presence of basement membrane proteins at the structurally and biochemically complex dermal-epidermal junction.Moreover, the homogeneous distribution of vimentin-positive cells in the dermis indicated the presence of NHDF.To define the cellular heterogeneity of the hdSE in more detail scRNAseq was performed using the droplet-based 10x Genomics platform 3 (Supplementary Fig. 3a).Data analysis of 5958 cells that passed quality control (Supplementary Fig. 3b and Supplementary Data 1) resulted in a tdistributed stochastic neighbor embedding (t-SNE) plot displaying 7 clusters with distinct expression profiles (Fig. 1d and Supplementary Data 1).All clusters were confirmed by the expression of known marker genes (Fig. 1e).Fibroblasts were identified in two clusters (clusters FB and FB-Pro) by their archetypal markers DCN and LUM 4 .SFRP2 and DPP4 expression was detected in the FB cluster, characterizing the most abundant major fibroblast population within the dermis.Cells expressing these markers are known to be small and elongated and express high levels of extracellular matrix components 5 .Keratinocytes were detected in 5 clusters and their diversity was mainly due to their degree of differentiation.Basal keratinocytes (clusters B-I, B-II, B-III, and B-Pro) highly expressed KRT5 and KRT14, whereas suprabasal keratinocytes (S) specifically expressed KRT1 and KRT10.Proliferating cells were present in both cell populations.They were identified by their expression of mitotic markers such as MKI67 and TOP2A (clusters FB-Pro and B-Pro).Differential gene expression analysis of the 5 keratinocyte subclusters revealed 3 clusters of basal keratinocytes (B-I, B-II, and B-III), which is in good agreement with native human skin 6 .The B-I cluster was, besides KRT5 and KRT14, also characterized by KRT15, DAPL1, and CXCL14.CXCL14 is constitutively expressed in skin and produced by keratinocytes 7 .Cluster B-II contained mainly genes that are important for the formation of hemidesmosomes and the attachment of the epidermis to the dermis via the basement membrane such as LAMA3, LAMB3, LAMC2, ITGA6, ITGB4, DST, PLEC, COL7A1, and COL17A1 (Fig. 1f, Supplementary Fig. 3c, and Supplementary Data 1).Cluster B-III was defined by RHCG, KRT6A, and KRT16, similar to what was found earlier, and it was assumed that these cells could directly differentiate into suprabasal cells 6 .Suprabasal cells in cluster S expressed known markers for epidermal differentiation such as SBSN or KRTDAP.Cluster B-Pro was enriched for the cell cycle markers UBE2C and PCNA.Furthermore, a detailed analysis of extracellular matrix (ECM)-associated genes revealed a remarkable repertoire of expressed ECM genes in the hdSE (Supplementary Fig. 3c).Many components of the ECM, including the collagens COL1A1, COL1A2, COL3A1, COL12A1, as well as elastin (ELN), fibronectin (FN1), and fibrillin (FBN1) were expressed in the fibroblast cluster.The same was true for factors involved in matrix assembly, such as MFAP4, SFRP2, LOX, FAP, and ANXA2, as well as for proteins involved in matrix remodeling, such as matrix metallopeptidases (MMP2, MMP14) and TIMP metallopeptidase inhibitors (TIMP1, TIMP3).Different culture media were tested to find optimal conditions for co-cultivation of parasites and skin equivalents (Supplementary Fig. 4a, b).Moreover, to test whether primary cells isolated from different human donors are compatible with each other, hdSEs consisting of NHDF and NHEK from three different donors were evaluated based on the thickness of the cellular epidermis (Supplementary Fig. 4b, c).Overall, these results indicate that trypanosomes and hdSEs can be cocultured with a mixture of trypanosome and skin medium without detrimental effects.

Supplementary Figure 1. Bioreactor design and compression system. (a) Representation of the computer-aided design (CAD) of the compression reactor with all individual parts. (b)
Cross section through the closed reactor after complete compression.The red arrows mark the route of the compressed air through the reactor.The compressed air was introduced through the compressed air connection, which could then exit through the channels in the individual compression punches to release the dermal equivalents.(c) Representation of the compression system.Shown is the configuration of the linear motor, which was attached to a stainless-steel motor mount via the motor flange.The motor can be controlled with the LinMot®-Talk software by connecting it to a computer.(d) Plot of oscillatory rheological measurements.Storage, G′ (open shapes), and loss modulus, G′′ (filled shapes), of cell-free and normal human dermal fibroblast (NHDF)-populated high-and lowdensity dermal equivalents as well as skin biopsies were determined in dependency on the strain deformation (ω = 6.28 rad/s, T = 22 °C).The linear viscoelastic region (gray box) was determined according to DIN 53019-4.Data represent means ± SD (low-density cell-free, n = 3; high-density cell-free, n = 3; high-density + NHDF, n = 6; skin biopsies, n = 7).Source data are provided as a Source Data file.(e) Assessment of NHDF viability in high-density dermal equivalents on day 23 after compression.NHDF were stained with fluorescein diacetate (FDA, cyan) and propidium iodide (PI, yellow) for live/dead discrimination.Scale bar, 100 µm.

Supplementary Figure 3. Strategy and quality control of the single-cell RNA sequencing experiment of high-density skin equivalents and analysis of the expression of extracellular matrix-associated genes. (a)
The epidermis and dermis of high-density skin equivalents were separated, cut, and dissociated by trypsin or liberase treatment.Cell suspensions were separated by fluorescence-activated cell sorting and single, live cells were sorted.Subsequently, cells were loaded on the droplet-based 10x platform according to the manufacturers' instructions and sequenced.(b) Quality control of 10x data showing the number of detected genes, UMIs (unique molecular identifiers), and the percentage of mitochondrial (mt) genes per cell in each of the 7 subclusters (fibroblasts (FB), proliferating fibroblasts (FB-Pro), basal (B-I -III), suprabasal (S), and proliferating keratinocytes (B-Pro)).The red lines in each graph represent quality control thresholds and cells above or below these thresholds were excluded from analysis.Due to the overall high number of genes identified to be expressed in BSF cells, the upper threshold was increased to < 2100 genes.(c) Feature plot of the 142 transcriptomes that passed quality control (excluding BSF) showing the normalized expression of fly-stage specific BARP genes.BARP-genes are specifically expressed in epimastigote forms and three such parasites were excluded from the MCF transcriptomes.The color key from blue to red indicates low to high gene expression levels.(d) Heatmap showing the scaled expression levels of upregulated genes of early (MCF + 4 h) and late (12 hpi + 24 hpi) timepoints, as identified by SCDE (absolute log2 fold change > 2, z-score > 1.96).Each column represents a single parasite, each row represents an individual gene.The color key from purple to yellow indicates low to high gene expression levels.(e) Volcano plot illustrating differential gene expression between early (MCF + 4 hpi) and late (12 hpi + 24 hpi) time points.Genes upregulated in early time points are depicted on the right side of the graph, genes upregulated in late time points on the left.x-and y-axis show log2 fold change (log2 FC) against the adjusted -log10 p-value (-log10P) and dashed lines indicate the set thresholds.Genes with absolute log2 FC > 2 and p-value < 0.01 are considered as differentially expressed.(f) Biological process-associated gene ontology (GO) terms significantly enriched in early (MCF + 4 h) and late (12 hpi + 24 hpi) timepoints.GO terms were filtered with the Revigo webtool to avoid redundancies.Bars represent the level of significance of the term enrichment determined by Fisher´s exact test.

Supplementary Figure 6 .
(c) Heatmap showing the scaled expression levels of extracellular matrix-associated genes differentially expressed in each cluster.The color key from pink to yellow indicates low to high gene expression levels.Each column represents a single cell, each row represents an individual gene.Gene names are listed to the left.assessment of the cell cycle distribution of skin-residing trypanosomes at various timepoints postinfection.(d) Flow cytometric profiles showing the gating strategy to detect trypanosomes (tdTomato positive) isolated from skin equivalents, to determine parasite viability (Calcein-AM positive), and to assess cell cycle distribution with NuclearGreen staining.Quality control of the single-parasite RNA sequencing experiment.(a, b) Violin and dot plots showing the number of genes detected across all libraries and library size of individual T. brucei parasites (metacyclic forms (MCF), n = 48; 4 h, n = 27; 12 h, n = 27; 24 h, n = 26; 7 d, n = 25; bloodstream forms (BSF), n = 31).The red lines represent the thresholds used for quality control.The BSF data set was prepared separately, indicated by the dashed line.