To gain a more comprehensive understanding of Bb gene expression throughout the tick phase of the transmission cycle, we developed an experimental approach to characterize the Bb transcriptome of spirochetes isolated from nymphal I. scapularis ticks during a days-long bloodmeal in which Bb is transmitted to a vertebrate host. We aimed to establish a longitudinal transcriptional profile encompassing key pathogen transmission events each day of feeding after ticks attached to their mouse bloodmeal hosts (Figure 1A). We fed Bb-infected nymphal ticks on naïve mice and collected the feeding ticks at daily intervals after the start of feeding until four days after attachment, at which time the ticks had fully engorged and detached from the mice. The major bottleneck for such an RNA sequencing (RNA-seq) approach is capturing sufficient quantities of Bb transcripts from complex multi-organism samples in which pathogen transcripts represent a very small minority. Our initial attempts to uncover Bb mRNA by simply removing tick mRNA with polyA-depletion and removing tick rRNA sequences using Depletion of Abundant Sequences by Hybridization (DASH) (Dynerman et al., 2020; Gu et al., 2016) were unsuccessful. This approach resulted in an average of only 0.09% of RNA-seq reads mapping to Bb mRNA – approximately 10-fold less than we estimated would be needed to feasibly obtain robust transcriptome-wide differential gene expression analysis (Haas et al., 2012).

To dramatically increase Bb transcript representation in our libraries, we physically enriched Bb cells in tick lysates prior to library preparation by adding an initial step of immunomagnetic separation (Figure 1B). We took advantage of a commercial antibody previously generated against whole Bb cells (αBb, RRID: AB_1016668). By western blot analysis, we confirmed that αBb specifically recognized several Bb proteins, including surface protein OspA (Figure S1A), which is highly prevalent on the Bb surface in the tick (Radolf et al., 2012). In addition, immunofluorescence microscopy with αBb showed clear recognition of Bb cells from within the tick at each day of feeding (Figure S1B). After collecting infected nymphal ticks from mice one, two, three, and four days post-attachment, we used αBb and magnetic beads to enrich Bb cells from the tick material in the lysates. We tracked relative Bb enrichment through RT-qPCR of Bb flaB RNA and tick gapdh RNA in the separated samples. Measuring Bb flaB RNA from both Bb-enriched samples and their matched Bb-depleted fractions, we found over 95% of total Bb flaB RNA was present in enriched fractions (Figure 1C), suggesting our approach captured the vast majority of Bb transcripts from the tick.

Total RNA recovered from the enrichment process was used to create RNA-seq libraries that subsequently underwent depletion of rRNA sequences from the tick, mouse, and Bb using DASH, which targets unwanted sequences for degradation by Cas9 (Gu et al., 2016; Ring et al., 2022). DASH reduced unwanted sequences from 94% to 9% of our total libraries, greatly increasing the relative abundance of Bb transcripts (Figure 1D). For resulting libraries generated across feeding, between 0.6% and 3.4% of reads mapped to Bb coding regions (Figure 1E), which translated to an average of 4.3 million Bb genic reads per sample (Figure 1F and Table S1).

As expected, for samples pulled from the mice one, two, and three days after attachment, the majority of the remaining sequencing reads mapped to the I. scapularis genome (72-83%) (Table S1). On day four, when ticks were fully engorged and were recovered from mouse cages rather than pulled from the mice, we found that fewer reads mapped to the I. scapularis genome (24-44%). Only a small percentage of reads from all samples mapped to the host Mus musculus genome (1-3%). To identify the source of the remaining reads in the day 4 samples, we ran the data through a publicly available computational pipeline that identifies microbes in sequencing datasets, CZ ID (Kalantar et al., 2020). This analysis led us to discover that a large percentage of day 4 reads mapped to bacterial species Pseudomonas fulva (41-64%) (Table S1), which may have been present in our mouse cages. While these samples had a lower percentage of reads that mapped to Bb, broad transcriptome coverage was still obtained by increasing total sequencing depth. Across all samples, at least 10 reads mapped to 92% of Bb genes. The median number of reads per gene in each sample varied from 338 to 1167 reads (Table S1). This coverage was sufficient for statistically significant downstream differential expression analyses for the vast majority of Bb genes.

To evaluate whether our approach introduced any major artifacts in Bb expression, we sequenced and compared RNA-seq libraries from in vitro cultured Bb cells before and after immunomagnetic enrichment. We found minimal expression differences (29 genes with p<0.05, fold changes between 0.83-1.12) (Figure S2 and Table S2), suggesting experimental enrichment did not significantly alter global transcriptome profiles for Bb. Thus, our enrichment approach enabled genome-wide analysis of Bb population-level expression changes that occur within the feeding nymph as Bb is transmitted to the host.

To provide a broad overview of Bb expression changes in the tick during the nymphal I. scapularis transmission bloodmeal, we performed principal component analysis (PCA) on the Bb transcriptome data from one, two, three, and four days after attachment (n=4). We reasoned that if many longitudinal expression changes were occurring across Bb populations, we would observe greater data variability between time points than between biological replicates. Indeed, we found replicates from each day grouped together, whereas distinct time points were largely non-overlapping. The first principal component, which explained 64% of the variance in our data, correlated well with day of feeding (Figure 2A). The global pattern suggested that Bb gene expression changes generally trended in the same direction over the course of feeding with the most dramatic differences between flanking timepoints on day 1 and day 4.

Using day 1 (early attachment) as a baseline, we performed differential expression analysis for all Bb genes at subsequent time points (day 2, day 3, day 4) (Figure 2B and Table S3). We examined changes with p-values < 0.05 when adjusted for multiple hypothesis testing and fold changes above a two-fold threshold (listed in Table S4). These analyses mirrored the global longitudinal expression pattern predicted by the PCA. The total number of differentially expressed (DE) genes when compared to day 1 increased with each subsequent timepoint to day 4. By day 4 there were 186 DE genes, including 153 upregulated and 33 downregulated (Figure 2C-E). Across all later time point comparisons to day 1, DE genes were highly overlapping and largely changed in the same directions. For example, of the DE genes that increased on day 2, 29 of 30 were still increased on day 3, and 29 of 30 were still increased on day 4. In the day 2, day 3, and day 4 comparisons to the day 1 baseline, we found 192 DE genes in total (Table S4). There were some notable differences between genes, such as the overall timing and kinetics of expression changes. Transcript levels for some DE genes changed suddenly over the course of feeding, while others were more gradual. To our knowledge, this is the first comprehensive report of global Bb expression changes over multiple stages of a tick feeding.

To assess the integrity of our dataset, we first examined expression profiles of previously characterized targets of major transcriptional programs activated at the onset of the bloodmeal: the RpoN/RpoS sigma factor cascade and the Hk1/Rrp1 two-component system (Caimano et al., 2015; Grassmann et al., 2023). Between day 1 and day 4, the expression of rpoS increased (Figure S3A), as expected (Hübner et al., 2001), and the majority of genes activated by RpoS in the feeding tick (Grassmann et al., 2023), including canonical targets ospC and dbpA, were also significantly upregulated (79/89, p<0.05, Wald tests) (Figure S3B). The majority of genes activated or repressed by Rrp1 in vitro (Caimano et al., 2015) also trended significantly in the expected direction ex vivo between day 1 and day 4 (111/148 upregulated, 37/57 downregulated, p<0.05, Wald tests, Figure S3C). We also examined the expression trends of genes regulated by RelBbu as part of the stringent response, another major transcriptional program active in Bb in the tick during nutrient starvation (Drecktrah et al., 2015). About half of RelBbu-regulated genes changed in the direction expected if the stringent response was active during this time (129/251 upregulated, 111/226 downregulated, p<0.05, Wald tests) (Figure S3D); however, more of the genes that were two-fold downregulated over feeding are regulated by RelBbu than either RpoS or Rrp1, suggesting it may play a role during this time frame.

We then compared the 192 two-fold DE genes in our longitudinal time course to those identified in two previous studies that measured Bb gene expression changes from culture conditions approximating the unfed tick and fed tick through modulation of temperature and/or pH (Ojaimi et al., 2003; Revel et al., 2002). 31% of the DE genes upregulated from day 1 (49/158) were more highly expressed in “fed tick” conditions compared to “unfed tick” conditions in one or both studies, while 24% of the DE genes downregulated from day 1 (8/24) were more highly expressed in “unfed tick” conditions in one or both studies (Figure S4A and Table S4). The studies become more concordant when focusing on the DE genes that were upregulated on day 2, which were generally the genes that changed the most dramatically in the time course. 70% of DE genes upregulated on day 2 (21/30) were more highly expressed in “fed tick” conditions in these previous studies, suggesting that the majority of the most dramatic gene expression changes we saw across feeding agree with what was observed in these previous studies.

We also compared the DE genes to two studies that assessed Bb gene expression differences in fed nymphs versus dialysis membrane chambers (DMCs), which mimic Bb conditions in the mammal (Grassmann et al., 2023; Iyer et al., 2015). 63% of all upregulated DE genes (100/158) were differentially expressed between fed nymphs and DMCs in one or both studies (Figure S4B and Table S4). The genes that were more highly expressed in nymphs were most concentrated amongst the day 2 DE genes (17/30, 57%), while the genes more highly expressed in DMCs were concentrated amongst the day 3 and day 4 DE genes (55/128, 43%). These comparisons suggested that the timing and magnitude of gene expression changes during feeding may indicate whether gene expression will peak in the tick or continue rising once Bb is transmitted to the host.

Through comparisons to these previous studies, we were able to verify that our data captured many expected transcriptional trends occurring during tick feeding. Nevertheless, 14% of the two-fold DE genes were not previously found to change expression in these different tick-feeding contexts (Grassmann et al., 2023; Iyer et al., 2015; Ojaimi et al., 2003; Revel et al., 2002) or identified in these RNA-seq studies as dependent upon RpoS, Rrp1, or RelBbu (Caimano et al., 2015; Drecktrah et al., 2015; Grassmann et al., 2023), which are the three main Bb regulatory programs active in the tick (Samuels et al., 2021). These additional genes highlight the necessity of measuring transcription in the tick environment and suggest we uncovered gene expression changes specific to the tick-stage of the Bb enzootic cycle. The nature and dynamics of these changes provide insights into potential genetic determinants of Bb survival, proliferation, and dissemination in the tick during transmission.

Bb has a complex, highly fragmented genome (Barbour, 1988) (Figure 3A), including numerous plasmids that are necessary during specific stages of the enzootic cycle (Schwartz et al., 2021) suggesting they contain genes that are crucial for pathogen transmission and survival. In fact, many genes found on the plasmids have been previously shown to alter expression upon environmental changes or in different host environments (Iyer et al., 2015; Ojaimi et al., 2005; Revel et al., 2002; Tokarz et al., 2004). Thus, we reasoned that many of the 192 DE Bb genes that change expression from day 1 to any later feeding time point (Table S4) would reside on the plasmids, and we examined their distribution throughout the genome. Consistent with these previous reports, we found that most of the upregulated genes were located on the plasmids (143/158; 90%), while fewer were found on the chromosome (15/158; 10%) (Figure 3B), which is home to the majority of metabolic and other housekeeping genes. In contrast, the majority of the downregulated genes were found on the chromosome (27/34, 79%) (Figure 3C).

Several plasmid-encoded genes that were longitudinally upregulated in our dataset have known roles during the tick bloodmeal or in mammalian infection. Linear plasmid 54 (lp54), which is an essential plasmid present in all Bb isolates (Casjens et al., 2012), contained the largest number of upregulated genes. Many of the genes on lp54 are known to be regulated by RpoS during feeding, including those encoding adhesins DbpA and DbpB, which are important for infectivity in the host (Blevins et al., 2008). This set also included five members of a paralogous family of outer surface lipoproteins BBA64, BBA65, BBA66, BBA71, and BBA73. BBA64 and BBA66 are necessary for optimal transmission via the tick bite (Gilmore et al., 2010; Patton et al., 2013). These findings indicate our dataset captures key Bb transcriptional responses known to be important for survival inside the tick during a bloodmeal.

Many upregulated genes were also encoded by cp32 plasmid prophages. Bb strain B31-S9 harbors seven cp32 isoforms that are highly similar to each other (Casjens et al., 2012). When cp32 prophages are induced, phage virions called ϕBB1 are produced (Eggers and Samuels, 1999). In addition to phage structural genes, cp32 contain loci that encode various families of paralogous outer surface proteins (Stevenson et al., 2000). Amongst the cp32 genes that increased over feeding were members of the RevA, Erp, and Mlp families, which are known to increase expression during the bloodmeal (Gilmore et al., 2001). We also found several phage genes that were upregulated, including those encoding proteins annotated as phage terminases on cp32-3, cp32-4, and cp32-7 (BBS45, BBR45, and BBO44). Some cp32 genes have been shown to change expression in response to the presence of blood (Tokarz et al., 2004), and as a part of the stringent response regulated by RelBbu (Drecktrah et al., 2015), while BBD18 and RpoS regulate prophage production in the tick midgut after feeding (Wachter et al., 2023). Our data suggest that some prophage genes are upregulated over the course of tick feeding, raising the possibility that cp32 prophage are induced towards the end of feeding. Overall, our data support the long-held idea that the Bb plasmids, which house many genes encoding cell envelope proteins, proteins of unknown function, and prophage genes, play a critical role in the enzootic cycle during the key transition period of tick feeding.

To gain a better overall sense of the types of genes that changed over feeding and the timing of those changes, we grouped DE genes into functional categories. Since a high proportion of plasmid genes encode lipoproteins within the unique protein-rich outer surface of Bb, genes of unknown function, and predicted prophage genes (Casjens et al., 2000; Fraser et al., 1997), we expected that many of the DE genes would fall into these categories. We classified the genes as related to either: cell envelope, bacteriophage, cell division, DNA replication and repair, chemotaxis and motility, metabolism, transporter proteins, transcription, translation, stress response, protein degradation, or unknown (Drecktrah et al., 2015) (see Table S4).

Of the genes that increased two-fold over feeding, the clear majority each day of feeding fell into two broad categories, cell envelope (55 of 158 total genes) and proteins of unknown function (69 of 158 total genes), with fewer genes related to metabolism, chemotaxis and motility, transporters, bacteriophage, cell division, and transcription (Figure 4A). In contrast to the upregulated genes, genes downregulated during feeding were more evenly distributed among the functional categories – including translation, protein degradation, transcription, and metabolism – consistent with many of them being located on the chromosome.

When looking at changes across these functional categories, the overrepresentation of cell envelope proteins was striking, while not unexpected. The Bb outer surface is covered with lipoproteins (Figure 4B), and these proteins are critical determinants in Bb interactions with the various environments encountered during the enzootic cycle (Kurokawa et al., 2020). We found that more than half (46 of 83) of annotated outer surface lipoproteins (Dowdell et al., 2017; Iyer et al., 2015) changed expression two-fold over the time course (Figure 4C). These data suggested widespread changes may be occurring on the Bb outer surface during feeding.

To understand the functional implications of expression changes in a majority of outer surface lipoproteins, we also compared their relative expression. The magnitude of expression varied greatly, with ospA, ospB, ospC, and bba59 being the most highly expressed outer surface protein transcripts. Many of the outer surface protein genes that we found increased expression over feeding were much less abundantly expressed (Figure 4C and Table S5). However, even the genes that appeared to have low expression in these population level measurements could play important roles in transmission if they are highly expressed in a small number of crucial cells, such as those that ultimately escape the midgut. While bulk RNA-seq cannot distinguish what is happening at the single cell level, our data suggest that during the bloodmeal, Bb are undergoing a complex outer surface transformation driven by increases in transcription of a majority of the genes encoding these lipoproteins.

Our RNA-seq data suggested that the outer surface of Bb transforms over the course of feeding as Bb are primed for transmission to a vertebrate host. At the same time, the tick midgut environment is changing as the tick begins to digest its bloodmeal (Sonenshine and Anderson, 2014). Tick-Bb interactions are likely crucial at the beginning of the tick bloodmeal, as Bb adhere to the tick gut epithelium before becoming motile and migrating out of the midgut and into the salivary glands (Dunham-Ems et al., 2009). Some Bb outer surface proteins, such as BBE31 and BBA52 play key roles in pathogen migration through interactions with the tick environment (reviewed in Kurokawa et al., 2020). We wanted to explore the changing tick environment to identify tick proteins with which Bb could interact throughout the tick bloodmeal. Since our Bb enrichment process retained some tick material, we reasoned that tick proteins that interact with Bb would be present in these samples.

To outline the changes occurring in the tick during feeding and to identify candidate Bb-interacting tick proteins, we used mass spectrometry to survey the content of the tick material that was enriched along with the Bb cells we sequenced at early and late stages during feeding. We purified proteins from the αBb-enriched fraction of crushed infected ticks one day after attachment and four days after attachment in triplicate. As controls for each day, we also performed the Bb enrichment process on lysate from uninfected ticks mixed with in vitro cultured Bb to help rule out proteins that were not pulled down through in vivo tick-Bb interactions (Figure 5A). When querying against the Bb and I. scapularis proteomes, we identified between 414 and 2240 protein groups per sample replicate. The vast majority of all detected proteins were from ticks (2801/2858, 98%). To identify proteins of interest, we looked for those that were detected in at least two of three replicates within infected ticks and had a mean average coverage twice that of uninfected ticks mixed with cultured Bb. We found 256 proteins (251 from I. scapularis and 5 from Bb) that were enriched with Bb from infected ticks one day after attachment and 226 proteins (220 from I. scapularis and 6 from Bb) that were enriched with Bb from infected ticks four days after attachment (Table S6). Of these proteins, only 27 (24 from I. scapularis and 3 from Bb) were detectable on both days, suggesting the tick proteins present upon Bb enrichment change dramatically over the course of feeding (Figure 5A). Amongst the small number of Bb proteins, we identified OspC in the samples from day 4 after attachment but not day 1 after attachment, confirming the expression change we saw in our RNA-seq. The distinct sets of tick proteins we identified at each timepoint suggest dramatic changes occur in the Bb-infected tick midgut environment during feeding that may alter the landscape of tick-Bb interactions.

Some of the proteins enriched with Bb from the tick may be good candidates for key Bb-interacting partners during feeding, especially if they are localized to the surface of tick cells where they may encounter Bb. The scarcity of both predicted and experimentally validated functions and localizations for tick proteins makes it difficult to fully assess the potential for tick protein interactions with extracellular Bb. Nevertheless, of the proteins found exclusively one day after attachment, 10 were categorized as extracellular matrix proteins using the PANTHER gene database (Thomas et al., 2022), and this category was statistically enriched (Fisher’s exact test, FDR=0.000093) (Figure 5B and Table S7). 30 additional proteins were annotated with a cellular component as plasma membrane. Four days after attachment, we did not detect any annotated extracellular matrix proteins; however, we identified 31 proteins that are likely to be found at the membrane, including two proteins annotated as putative low-density lipoprotein receptors (Figure 5C and Table S8). These extracellular matrix and membrane proteins may be the most likely to directly interact with Bb during this timeframe and are candidates for tick proteins important in the Bb dissemination process. The proteins present in the changing tick environment may be key determinants of pathogen transmission as Bb remodels its outer surface while preparing to migrate through the tick to a new host.

Bb strain B31-S9 (Rego et al., 2011) was provided by Dr. Patricia Rosa (NIAID, NIH, RML) and cultured in BSK II media at 35°C, 2.5% CO2. B31-S9 was used for all RNA-seq and Bb enrichment experiments. Wildtype Bb strain B31-A3, ospA1-mutants (ospA1) and ospA-restored Bb (ospA⁺B1) (Battisti et al., 2008) used in αBb western blot were also provided by Dr. Rosa.

I. scapularis larvae were purchased from the Tick Lab at Oklahoma State University (OSU) for RNA-seq experiments or provided by BEI Resources, a division of the Center for Disease Control, for mass spectrometry experiments. Before and after feeding, ticks were maintained in glass jars with a relative humidity of 95% (saturated solution of potassium nitrate) in a sealed incubator at 22°C with a light cycle of 16h/8h (light/dark). Animal experiments were conducted in accordance with the approval of the Institutional Animal Care and Use Committee (IACUC) at UCSF, Project Number AN183452. Ticks were fed on young (4–6-week-old) female C3H/HeJ mice acquired from Jackson Laboratories. Mice were anesthetized with ketamine/xylazine before placement of ≤ 100 larval or ≤ 30 nymphal ticks. Replete larval ticks were placed in the incubator to molt before being used as nymphs in experiments. Nymphal ticks were either pulled off isoflurane anesthetized mice at various times during feeding (1-3 days after placement) or allowed to feed to repletion and collected from mouse cages (4 days after placement).

To determine whether the αBb antibody targeted ospA, wildtype Bb (B31-A3), ospA1-mutants (ospA1) and ospA-restored Bb (ospA⁺B1) (Battisti et al., 2008) were cultured to approximately 5×10⁷ Bb/mL. 3 mLs of culture were centrifuged for 7 minutes at 8000 x g, washing twice with PBS. Pelleted cells were lysed in 50 µL of water, and 25 µg of protein per sample were mixed with 5X loading dye (0.25% Bromophenol Blue, 50% Glycerol, 10% Sodium Dodecyl Sulfate, 0.25M Tris-Cl pH 6.8, 10% B-Mercaptoethanol), run on a Mini-PROTEAN TGX 4-15% gel (Biorad), and transferred using the Trans-Blot Turbo Transfer System (Biorad). After transfer, the blot was blocked for 30 minutes at 4°C in TBST (Tris buffered saline with 0.1% tween) with 5% milk, then treated with αBb antibody (Invitrogen: PA1-73004; RRID: AB_1016668) diluted 1:10,000 for 1 hour at room temperature, followed by anti-rabbit HRP secondary antibody (Advansta: R-05072-500; RRID: AB_10719218) diluted 1:5,000 for 45 minutes at room temperature with 3 short PBST washes between each step. Blots were exposed using Clarity Western ECL Substrate (Biorad) and imaged using the Azure C400 imaging system (Azure Biosystems). This experiment was repeated three times.

To sequence RNA from Bb from feeding ticks, Bb were enriched to increase the ratio of Bb to tick material. Larval ticks were fed to repletion on three mice that were infected with Bb through intraperitoneal and subcutaneous injection with 10⁴ total Bb. Approximately five months later, the molted nymphal ticks were fed on eight mice, which were housed individually during the feeding. We estimated that 83% of the ticks were infected with Bb by crushing 12 unfed nymphs in BSK II media and checking for viable Bb days later. Ticks were pulled from all mice and pooled into four biological replicates 1 day after placement (14 ticks per replicate), 2 days after placement (12 ticks per replicate), and 3 days after placement (6 ticks per replicate) and collected from cages 4 days after placement (7 ticks per replicate). Shortly after collection, ticks were washed with water and placed in a 2 mL glass dounce grinder (Kimble) in 500 µL of phosphate-buffered saline (PBS). Ticks were homogenized first with the large clearance pestle and then the small clearance pestle. The homogenate was transferred to a 1.5 mL Eppendorf tube and 500 µL of PBS was added to total 1 mL. At this stage, 50 µL of homogenate was removed as an input sample and mixed with 500 µL of TRIzol (Invitrogen) for RNA extraction. 2 µL of αBb antibody (Invitrogen: PA1-73004; RRID: AB_1016668) was added to the homogenate, which was then placed on a nutator at 4°C for 30 minutes. During incubation, 50 µL of Dynabeads Protein G (Invitrogen) per sample were washed twice in PBS. After incubation with the antibody, the homogenate and antibody mixture were added to the beads. This mixture was placed on a nutator at 4°C for 30 minutes. Tubes were then placed on a magnet to secure beads, and the homogenate was removed and saved to create depleted samples. The depleted homogenate was centrifuged at 8,000 x g for 7 minutes, 900 µL of supernatant was removed, and 500 µL of TRIzol was added to the pellet to create depleted samples. The beads were washed twice with 1 mL of PBS, resuspending the beads each time. The second wash was removed and 500 µL of TRIzol was added to the beads to create enriched samples. RNA was extracted from all input, enriched, and depleted samples using the Zymo Direct-zol RNA Microprep Kit with on-column DNase treatment (Zymo Research). The step-by-step Bb enrichment protocol is available at: dx.doi.org/10.17504/protocols.io.36wgqjrbovk5/v1.

To test whether the Bb enrichment process altered gene expression levels, we performed the enrichment protocol on cultured Bb. Tubes of Bb in BSK II media were grown to 9×10⁴ Bb/mL at 35°C. 1 mL of culture was spun down at 8,000 x g for 7 minutes, media was removed, Bb were washed in 1mL of PBS and spun again. Pelleted Bb were resuspended in 1 mL of fresh PBS. These samples were used as starting homogenate for the Bb enrichment protocol and input, enriched, and depleted fractions were collected as above. RNA-seq libraries from these samples were prepared and sequenced as below.

To make RNA-seq libraries from enriched Bb RNA, 50 ng of total RNA was used as input into the NEBNext Ultra II Directional RNA Library Prep Kit for Illumina (New England BioLabs). Libraries were prepared following the manufacturer’s protocol for use of the kit with purified mRNA or rRNA-depleted RNA, despite starting with total RNA. Libraries were barcoded using NEBNext Multiplex Oligos for Illumina Dual Index (New England BioLabs).

To deplete reads from the libraries that were from tick, Bb, or mouse rRNA, we used Depletion of Abundant Sequences by Hybridization (DASH) (Gu et al., 2016), which targets Cas9 to unwanted reads in RNA-seq libraries using custom dual-guide RNAs (dgRNAs). The dgRNAs targeted short sequences within tick rRNA, mouse rRNA, and Bb rRNA that were designed using DASHit software (Dynerman et al., 2020). We ordered crRNA oligos (Table S9) that targeted these sequences (Ring et al., 2022). To transcribe the dgRNAs, we followed the protocol for In Vitro Transcription for dgRNA V2 (Lyden et al., 2019b) as follows. Both tracrRNA and pooled crRNA DNA templates were annealed to equimolar amounts of T7 primer by heating to 95°C for 2 minutes and slowly cooling to room temperature. Annealed templates were used in 1 mL in vitro transcription reactions with: 120 μL10X T7 buffer (400 mM Tris pH 7.9, 200 mM MgCl2, 50 mM DTT, 20 mM spermidine (Sigma)), 100 μL of T7 enzyme (custom prepped enzyme from E. Crawford, diluted 1:100 in T7 buffer, final concentration: 100 μg/mL), 300 μl NTPs (25 mM each, Thermo Fisher Scientific), 4 μg of annealed crRNA template or 8 μg of annealed tracrRNA template, and water to 1 mL. In vitro transcription was performed for 2 hours at 37°C. Reactions were purified twice with the Zymo RNA Clean & Concentrator-5 Kit (Zymo Research). To form the dgRNA complex for DASH, crRNA and tracrRNAs were diluted to 80 μM, mixed in equimolar amounts and annealed by heating to 95°C for 30 seconds and cooling slowly to room temperature.

After transcription of dgRNAs, we performed DASH Protocol Version 4 (Lyden et al., 2019a) on each library individually as follows. Cas9 and transcribed dgRNAs were prepped by mixing: 2.5 μL 10X Cas9 buffer, 5 μL 20 μM Cas9 (New England BioLabs), and 5 μL of 40 μM transcribed dgRNAs. The mixture was incubated at 37°C for 5 minutes before 7.5 μL of RNA-seq library (2.8 nM) was added. The mixture was incubated at 37°C for 1 hour and then purified with Zymo DNA Clean & Concentrator-5 (Zymo Research) following the PCR product protocol and eluting DNA into 10.5 μL of water. During cleanup, Cas9 was again mixed with buffer and dgRNAs and incubated at 37°C for 5 minutes. Following cleanup, eluted DNA was added to the second Cas9-dgRNA mixture and incubated at 37°C for 1 hour for a second time. Then, 1 μL of proteinase K (New England Biolabs) was added, and the mixture was incubated at 50°C for 15 min. The libraries were then purified with 0.9X volume of sparQ PureMag Beads (QuantaBio) following the standard protocol, eluting in 24 μL of water. rRNA-depleted RNA-seq libraries were then amplified in a BioRad CFX96 using the Kapa HiFi Real-Time Amplification Kit (Roche) in a 50 μL reaction with 25 μL master mix, 23 μL of the DASHed library pool, and 2 μL of a 25 μM mix of Illumina P5 (5’-AATGATACGGCGACCACCGAGATCT) and P7 (5’-CAAGCAGAAGACGGCATACGAGAT) primers. The qPCR program for amplification was as follows: 98°C for 45 sec (1 cycle), (98°C for 15 sec, 63°C for 30 sec, 72°C for 45 sec, plate read, 72°C for 20 sec) for 10 cycles (day 3 and day 4 samples) or 11 cycles (day 1 and day 2 samples). The libraries were removed from cycling conditions before leaving the exponential phase of amplification and then purified with 0.9X volume of sparQ PureMag Beads according to the standard protocol.

Following DASH, RNA-seq libraries were sequenced on an Illumina NovaSeq S2 (2 lanes) with paired-end 100 base pair reads. Libraries from in vitro cultured control experiment were sequenced on an Illumina NextSeq with paired-end 75 base pair reads. FASTQ files and raw Bb read counts for in vitro control experiment (GSE217146) and ex vivo experiment (GSE216261) have been deposited in NCBI’s Gene Expression Omnibus (Edgar et al., 2002) under SuperSeries accession number GSE217236.

To test the efficacy of the Bb enrichment protocol, RT-qPCR was used to quantify Bb flaB and I. scapularis gapdh transcript levels in enriched and depleted fractions. cDNA was synthesized from 8 μL of RNA extracted from Bb enrichment samples and their matched depleted samples from day 2, day 3, and day 4 post-attachment using the qScript cDNA Synthesis Kit (Quantabio). cDNA was diluted 2X before use in qPCR. To measure flaB copies, standards of known concentration were created from purified PCR products. These standards were made from PCR with primers with the following sequences: 5’-CACATATTCAGATGCAGACAGAGGTTCTA and 5’-GAAGGTGCTGTAGCAGGTGCTGGCTGT. A dilution series with ten-fold dilutions between 10⁶ copies and 10¹ copies of this PCR template was run alongside enriched and depleted samples. qPCR was performed using Taqman Universal PCR Master Mix (Applied Biosystems). The primers used to amplify flaB were: 5’-TCTTTTCTCTGGTGAGGGAGCT and 5’-TCCTTCCTGTTGAACACCCTCT (used at 900 nM) and the probe was /56-FAM/AAACTGCTCAGGCTGCACCGGTTC/36-TAMSp (used at 250 nM). For tick gapdh RT-qPCR, the cDNA samples were diluted an additional 2X. Standards of known concentration were created using the qPCR primer sequences: 5’-TTCATTGGAGACACCCACAG and 5’-CGTTGTCGTACCACGAGATAA (used at 900 nM). qPCR was performed using PowerUp SYBR Green Master Mix (Applied Biosystems). For both flaB and gapdh, the number of copies in each sample was calculated based on the standards of known concentration. Three technical replicates were averaged from each of four biological replicates at each time point tested. We totaled the number of copies in each matched enriched and depleted fraction to calculate the percentage of flaB or gapdh that was found in either sample. All qPCR was performed on the QuantStudio3 Real-Time PCR System (Applied Biosystems). Paired t tests were performed using GraphPad Prism v9.5.1.

To test whether the αBb antibody recognized Bb inside of the tick, ticks at each day of feeding were crushed in 50 µL of PBS. 10 µL of lysate was spotted onto slides and allowed to air dry before slides were heated briefly three times over a flame. Heat fixed slides were then treated with acetone for one hour. Slides were incubated with αBb primary antibody (1:100 diluted in PBS + 0.75% BSA) for 30 minutes at 37°C in a humid chamber. A control without primary antibody was also used for each day. Slides were washed once in PBS for 15 minutes at room temperature, then rinsed in distilled water and air dried. Anti-rabbit IgG Alexa 488 (Invitrogen: A-11008; RRID: AB_143165) diluted 1:100 in PBS + 0.75% BSA was added for 30 minutes at 37°C in a humid slide chamber. Slides were washed in PBS for 15 minutes at room temperature three times, adding 1:100 Propidium Iodide (Invitrogen) during the second wash. Slides were then rinsed with distilled water and air dried before the addition of mounting media (Fluoromount-G, SouthernBiotech) and cover slips. Fluorescence imaging was performed on a Nikon Ti2 inverted microscope for widefield epifluorescence using a 100X/1.40 objective. Images were captured with NIS-Elements AR View 5.20 and then processed with ImageJ software (Schneider et al., 2012). No strong florescence signal was observed on the control slides without primary antibody.

To identify which tick proteins were found in samples after Bb enrichment across feeding, both uninfected and infected ticks were fed on mice. Three biological replicates of uninfected ticks one day after attachment (11 ticks per replicate), infected ticks one day after attachment (27 ticks per replicate), uninfected ticks four days after attachment (8 ticks per replicate), and infected ticks four days after attachment (16 ticks per replicate) were collected. Before αBb enrichment, the uninfected tick samples were mixed with Bb grown in culture that was washed with PBS (3 x10⁴ Bb one day after attachment and 3×10⁶ Bb four days after attachment) and mixed lysates were rotated at room temperature for 30 minutes. Infected tick samples underwent the Bb enrichment process immediately. The enrichment process followed the same protocol used for RNA-seq. Sample volumes were increased to 1 mL as needed, and then 2 µL of αBb antibody (Invitrogen: PA1-73004; RRID: AB_1016668) was added, and samples were rotated at 4°C for 30 minutes. 50 µL of Dynabeads Protein G per sample were washed in PBS during this incubation and added to the lysates, which were rotated at 4°C for 30 minutes. The beads were washed twice with 1 mL of PBS, and then placed into 50 µL of lysis buffer (iST LYSE, PreOmics). Samples were boiled at 95°C for 5 minutes, and lysates were removed from beads and frozen for mass spectrometry preparation.

For mass spectrometry, a nanoElute was attached in line to a timsTOF Pro equipped with a CaptiveSpray Source (Bruker). Chromatography was conducted at 40°C through a 25 cm reversed-phase C18 column (PepSep) at a constant flowrate of 0.5 μL min−1. Mobile phase A was 98/2/0.1% water/MeCN/formic acid (v/v/v) and phase B was MeCN with 0.1% formic acid (v/v). During a 108 min method, peptides were separated by a 3-step linear gradient (5% to 30% B over 90 min, 30% to 35% B over 10 min, 35% to 95% B over 4 min) followed by a 4 min isocratic flush at 95% for 4 min before washing and a return to low organic conditions. Experiments were run as data-dependent acquisitions with ion mobility activated in PASEF mode. MS and MS/MS spectra were collected with m/z 100 to 1700 and ions with z = +1 were excluded.

Raw data files were searched using PEAKS Online Xpro 1.6 (Bioinformatics Solutions Inc.). The precursor mass error tolerance and fragment mass error tolerance were set to 20 ppm and 0.03 respectively. The trypsin digest mode was set to semi-specific and missed cleavages was set to 2. The I. scapularis reference proteome (Proteome ID UP000001555, taxon 6945) and Bb reference proteome (Proteome ID UP000001807, strain ATCC 35210/B31) was downloaded from Uniprot, totaling 21,774 entries. The I. scapularis proteome was the primary search reference and the Bb was used as a secondary to identify any bacterial proteins present. Carbamidomethylation was selected as a fixed modification. Oxidation (M) and Deamidation (NQ) was selected as a variable modification.

Experiments were performed in biological triplicate, with samples being a single run on the instrument. Proteins present in a database search (−10 log(p-value) ≥ 20, 1% peptide and protein FDR) were subjected to the following filtration process. Proteins were filtered to include only those found in 2 out of 3 biological replicates, within each respective day (one or four days after attachment). The mean area of proteins found in uninfected and infected samples was calculated. Proteins with missing values (i.e. not identified in a sample) were set to 1. The ratio of mean area for each protein was calculated as infected/uninfected, and enriched proteins were identified by having an infected to uninfected ratio greater than 2 within their respective feeding day (one or four days after attachment).

To classify the identified proteins into functional groups, a PANTHER Overrepresentation Test (released 07/12/2022) was used (Mi et al., 2019) with PANTHER version 17.0 (released 02/22/2022) (Thomas et al., 2022). All I. scapularis genes in the database were used as the reference list to and all proteins enriched on each day were analyzed for their PANTHER Protein Class. The test type used was Fisher’s Exact, calculating a false discovery rate as the correction. Raw data files and searched datasets are available on the Mass Spectrometry Interactive Virtual Environment (MassIVE), a full member of the Proteome Xchange consortium under the identifier: MSV000090560.