A baseline transcriptional signature associates with clinical malaria risk in RTS,S/AS01-vaccinated African children

In a phase 3 trial in African infants/children, the RTS,S/AS01 (GSK) vaccine showed moderate efficacy against clinical malaria. We aimed to identify RTS,S/AS01-induced signatures associated with clinical malaria by analyzing antigen-stimulated peripheral blood mononuclear cells sampled from a subset of trial participants at baseline and month 3 (one month post-third dose). RTS,S/AS01 vaccination was associated with downregulation of B-cell and monocyte-related blood transcriptional modules (BTMs) and upregulation of T-cell related BTMs, as well as higher month 3 (vs baseline) circumsporozoite protein (CSP)-specific CD4+ T-cell responses. There were few RTS,S/AS01-associated BTMs whose month 3 levels correlated with malaria risk. In contrast, baseline levels of BTMs associated with dendritic cells and with monocytes (among others) correlated with malaria risk. A cross-study analysis supported generalizability of the baseline dendritic cell- and monocyte-related BTM correlations with malaria risk to healthy, malaria-naive adults, suggesting inflammatory monocytes may inhibit protective RTS,S/AS01-induced responses.

NANP-specific, HBS-specific, and C terminal domain of CSP (C-term)-specific antibody data from 180 previous studies were analyzed for correlations with BTM expression as described below. IgG titers 181 (EU/mL) against NANP and against HBS were obtained from the MAL055 trial database (8)(9)(10)(11). IgG 182 titers (EU/mL) against NANP and C-term were measured by ELISA at IAVI-HIL (21). IgG and IgM 183 levels (Median Fluorescence Intensity, MFI) against NANP, C-terminal CSP and HBS together with 35 184 RTS,S/AS01 vaccine-unrelated malaria antigens were measured by Luminex technology (22,23). Preprocessing: Preprocessing of RNAseq data was done by Broad Technology Labs. In brief, reads 193 were aligned using BWA Aln version 0.7.10 using UCSC RefSeq (Human 19) with mitochondrial 194 genes added. Quantified samples were then quality controlled using mapping summary statistics to 195 remove low quality samples based on predetermined minimum values for the total number of mapped 196 reads, percent of mapped reads mapped to the human genome, etc. Downstream analysis was applied 197 only to reads that mapped uniquely to a UMI and only mapped to isoforms of the same gene 198 (UMI.unq). 199 200 Normalization: The TMM normalization method (29) was applied to account for differing number of 201 read counts and to address unwanted technical variation. The voom transformation (30) from the limma 202 R package (31) was applied to standardize and appropriately weight the data for use in linear models. 203 204 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted May 19, 2021. ;https://doi.org/10.1101https://doi.org/10. /2021 Quality control: In a pilot study, we found that sample libraries that exhibit less than 75,000 total 205 RNAseq reads per sample were of low quality. Thus, such libraries were removed from the study. 206 Genes that had less than 20 samples (around 10%) with read counts greater than 5 were also removed. 207 For further details, see the Supplementary Methods. 208 209 Differential expression: Differential expression was assessed using module-based (using voom and 210 camera (32)) approaches as implemented in the limma package. Camera, combined with voom, is one 211 of the few gene set enrichment analysis methods that can properly account for inter-gene correlation in 212 RNA-seq data. Specifically, Camera estimates the variance inflation factor for the gene expression that 213 results from inter-gene correlation in the data and incorporates it into test procedures to control the 214 apparent false discovery rate. This step is important since significant correlation is expected among 215 genes in the same module. Inference was based on p-values adjusted for multiple testing by controlling 216 the false discovery rate with the   For each module, a score was calculated for each RTS,S recipient at month 3 and at month 0 based on 228 the average normalized expression level of all genes in the modules, on the log scale. Spearman's rank 229 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted May 19, 2021. ;https://doi.org/10.1101https://doi.org/10. /2021 correlation was used to assess association between gene expression, functional antibody and cellular 230 responses. Each correlation was tested (Pearson correlation test) and a p-value was obtained. P-values 231 were adjusted within each response (across all gene sets); significance was defined as an adjusted p-232 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted May 19, 2021. ;https://doi.org/10.1101https://doi.org/10. /2021

BTMs, along with upregulation of T cell-related BTMs 277
The transcriptional response to RTS,S/AS01 vaccination was assessed in control-stimulated PBMC as 278 well as Ag-stimulated PBMC. Through this approach, we hypothesized that we would see recall 279 responses of Ag-specific T cells activated in vitro, as well as responses of other cell types to the 280 secreted cytokines/chemokines. Of note, the sampling schedule at MAL067 was designed for 281 evaluation of acquired immune responses to the vaccine and not ex vivo responses. Our motivation was 282 that in healthy, malaria-naïve adults, the transcriptional response to RTS,S/AS01 has been shown to 283 largely wane by Week 3 post-final dose (16), implying that the majority of the RTS,S/AS01-induced 284 transcriptional changes in this study likely preceded the month 3 sample collection. Three antigens 285 were chosen for stimulation: CSP (peptides covering the CSP region of RTS,S that encodes B-cell and 286 T-cell epitopes), HBS (peptides covering the HBS, also included in the RTS,S vaccine), and AMA1 (a 287 highly immunogenic antigen expressed briefly on hepatocyte-invading P. falciparum sporozoites and 288 predominantly on red blood cell-invading P. falciparum merozoites, not present in the RTS,S vaccine; 289 included to analyze naturally acquired immunity responses). 290 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted May 19, 2021. ;https://doi.org/10.1101https://doi.org/10. /2021 Two comparisons were done to characterize the transcriptional response to RTS,S/AS01 vaccination: 291 Comparison (1): comparing gene expression in month 3 samples from RTS,S/AS01 vs comparator 292 recipients (month 3 RTS,S/AS01 vs comparator); and Comparison (2): comparing gene expression in 293 month 3 vs month 0 from RTS,S/AS01 recipients (RTS,S/AS01 month 3 vs month 0). Each 294 comparison has its own advantages: Comparison (1) allows the identification of RTS,S/AS01-specific 295 responses while taking into account other environmental factors to which the children are exposed, 296 such as malaria exposure (albeit malaria transmission intensity was low during the study at both sites).  Table S3a). 308 The vast majority (55) of these BTMs were in vehicle-treated PBMCs, with few differences observed 309 in antigen-stimulated PBMCs (0 significantly differently expressed BTMs in CSP-stimulated PBMC 310 and 8 significantly differentially expressed BTMs in AMA1-stimulated PBMC). The categories with 311 the most differentially expressed BTMs were B cells and T cells: In vehicle-stimulated PBMCs from 312 infants/children who received RTS,S/AS01, 11 T cell-related BTMs were upregulated (vs comparator). 313 We also observed that natural killer (NK) cell and mitochondria-related BTMs were upregulated in 314 vehicle-stimulated PBMC, whereas B cell-related BTMs were downregulated. Counter to our initial 315 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted May 19, 2021. ;https://doi.org/10.1101https://doi.org/10. /2021 expectations, no significant correlations were identified in the CSP-stimulated cells. This result may 316 potentially be explained by the low frequency of CSP-specific T-cells in RTS,S/AS01 vaccinees (e.g. 317 on average, <0.10% of all CD4+ T cells (28)). In AMA1-stimulated cells, surprisingly, the few 318 correlate BTMs that were identified tended to correlate in opposite directions with risk than in the 319 vehicle-stimulated PBMC (i.e. a positive correlate in vehicle-stimulated PBMC yet an inverse correlate 320 in AMA1-stimulated PBMC, and vice versa). We hypothesize that cytokines/chemokines released from 321 activated T cells and their effects on other PBMC may underlie this difference. Alternatively, AMA1 322 may be eliciting an innate response, which is supported by the findings of Bueno et al. in Table S3b). 341 Antiviral/interferon (IFN)-, inflammatory/Toll-like receptor (TLR)/chemokine-, dendritic cell-, and 342 transcription-related BTMs were among those upregulated, whereas B cell-and monocyte-related 343 BTMs were among those downregulated. 344 The two Comparisons yielded partially divergent results, likely due to the reasons stated above. 345 However, downregulation of monocyte-related BTMs in vehicle-treated PBMC was consistently 346 observed across the two Comparisons: "Enriched in monocytes (II) (M11.0)" was downregulated in 347 both Comparisons, "Enriched in myeloid cells and monocytes (M81)" was downregulated in 348 Comparison (1), and "Monocyte surface signature (S4)" and "Enriched in monocytes (surface) 349 (M118.1)" were downregulated in Comparison (2). Interestingly, monocyte-related BTMs (including 350 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted May 19, 2021. ; https://doi.org/10.1101/2021.05.19.21257227 doi: medRxiv preprint these four BTMs) were upregulated at multiple time-points in response to RTS,S/AS01 administration 351 in (16), pointing to differences between the two study populations. . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted May 19, 2021. ; https://doi.org/10.1101/2021.05.19.21257227 doi: medRxiv preprint 383 Antigen-stimulated PBMC also showed transcriptional modules associated with clinical malaria risk. 384 However, these results differed largely from those seen in vehicle-stimulated PBMC. The two 385 monocyte-related BTMs (M81, M11.0) that were directly associated with risk in vehicle-stimulated 386 PBMC were inverse correlates in both HBS-and AMA1-stimulated PBMC. Moreover, five T-cell 387 related BTMs were positive correlates of risk in AMA1-stimulated PBMC (Figure 3). 388 The same analysis was performed on comparator recipients ( Figure S1). For all five BTMs whose In addition to transcriptional changes, our group has shown previously that RTS,S/AS01 vaccination 397 elicits vaccine-specific antibody and cellular responses in African infants and children (e.g. (21, 23, 398 28)). The polyfunctionality score is a summary measure that encapsulates a participant's entire Ag-399 specific T-cell response after vaccination (39). Using data from a pilot study of 179 children (none of 400 whom was a malaria case) at the Manhiça and Bagamoyo sites, Moncunill et al. previously showed that 401 MAL067 RTS,S/AS01 recipients have higher month 3 CSP-specific and HBS-specific CD4+ T-cell 402 polyfunctionality scores than comparator recipients (28). Consistent with this finding, we report that 403 average CSP-specific CD4 + T-cell polyfunctionality and magnitude (frequency of CD4 + T-cell 404 expressing IL-2 or TNF or CD154) are both higher at month 3 vs. baseline in RTS,S/AS01 vaccine 405 recipients ( Figure 4). The few high responders at baseline can likely be attributed to prior malaria 406 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted May 19, 2021. ; https://doi.org/10.1101/2021.05.19.21257227 doi: medRxiv preprint exposure. However, there was no difference in average month 3 CSP-specific T-cell response 407 polyfunctionality or magnitude between RTS,S/AS01 cases vs controls ( Figure S2  CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted May 19, 2021. ; https://doi.org/10.1101/2021.05.19.21257227 doi: medRxiv preprint 5, Table S5). The three antibody variables whose month 3 levels had the strongest positive correlations 425 with transcriptional data were "IgM, LSA1", "IgM, MSP1 b12 Mad20", and "IgM, MSP6", which 426 tended to be positively correlated with month 3 levels of DC-, inflammatory/TLR/chemokine-, and 427 monocyte-related BTMs. In contrast, month 3 levels of "IgG, AMA1 3D7" and "IgG, AMA1 FVO" 428 tended to be inversely correlated with month 3 levels of DC-and monocyte-related BTMs. None of 429 these associations were seen in comparator recipients ( Figure S3), suggesting specificity to 430 RTS,S/AS01 receipt, although we note that sample size is smaller which would have reduced statistical 431 power to detect differences. Month 3 levels of cellular variables assessed by polychromatic flow 432 cytometry did not correlate significantly with the month 3 level of any BTM. 433 434 435 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)   (24)). 455 In general, much fewer associations were seen for comparator recipients ( Figure S3). Specifically, the 456 inverse associations of CSP-specific IFN-γ responses with month 3 levels of DC-related BTMs were 457 not seen, suggesting that these associations are RTS,S/AS01-specific. 458 459 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

vaccinees between baseline expression of DC-and monocyte-related BTMs and risk 461
An important question is whether the results of our analysis of the MAL067 trial, which was conducted 462 in African infants and children in malaria-endemic areas, are generally translatable to other study 463 populations. PBMC transcriptomic data are available for at least three different controlled human 464 malaria infection (CHMI) studies conducted in a quite different study population, i.e. healthy, malaria-465 naïve adults in the United States. WRAIR 1032 (NCT00075049) randomly assigned participants to 466 receive RTS,S/AS02A or RTS,S/AS01B at months 0, 1, and 2 (40); MAL068 (NCT01366534) 467 randomly assigned participants to receive Ad35.CS.01 at month 0 followed by RTS,S/AS01B at 468 months 1 and 2 (heterologous prime-boost) or RTS,S/AS01B at months 0, 1, and 2 (14); and MAL071 469 (NCT01857869) randomly assigned participants to receive a full dose of RTS,S/AS01B at months 0, 1, 470 and 2 or a full dose of RTS,S/AS01B at months 0 and 1, followed by a fractional dose at month 7 (42). 471 Importantly, all these trials share a common vaccine arm: one full dose of RTS,S/AS01B at months 0, 472 1, and 2 (referred to as the "RRR" arm). Microarray data from WRAIR 1032 were analyzed by Vahey 473 et al. (43), microarray data from MAL068 were analyzed by Kazmin et al. (16), and RNA-seq data 474 from MAL068 and MAL071 were analyzed by Du et al. (17). 475 We performed a cross-study immune correlates analysis where we examined whether the BTMs 476 associated with clinical malaria risk in MAL067 showed similar associations with challenge outcome 477 in each of the three CHMI studies described above. Due to differences in sampling schedules, and the 478 presence of the CHMI challenge (which would complicate results interpretation), we could not 479 compare the exact same month 3 timepoint across studies. We chose instead to compare 21 days post-480 third dose in MAL068 and in MAL071, i.e. of day of challenge, and 14 days post-third dose in WRAIR 481 1032, i.e. just before or on day of challenge. We refer to these slightly different post-vaccination time 482 points as "month 3" for simplicity. The month 3 cross-study correlates analysis included BTMs whose 483 month 3 levels (in vehicle-stimulated PBMC) associated with clinical malaria risk in MAL067 484 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted May 19, 2021. ; https://doi.org/10.1101/2021.05.19.21257227 doi: medRxiv preprint RTS,S/AS01B recipients (Figure 3, Table S4) and is shown in Figure 6A. No BTM was consistently 485 associated with malaria risk (or non-protection) across all four studies. The most consistent result was 486 for the monocyte-related BTMs, where month 3 expression of "enriched in monocytes (II) (M11.0)" 487 was significantly associated with risk in 3 of the 4 studies (all except WRAIR 1032), and "enriched in 488 myeloid cells and monocytes (M81)" was significantly associated with risk in MAL067 and MAL071 489 RRR ( Figure 6A, Table S6A). 490 491 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted May 19, 2021. is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted May 19, 2021. ; https://doi.org/10.1101/2021.05.19.21257227 doi: medRxiv preprint difference in month 3 PBMC BTM expression between RTS,S/AS01 malaria cases vs. non-malaria 497 controls in each of three CHMI studies of the 5 BTMs whose month 3 levels in DMSO-stimulated 498 PBMC associated with malaria case status in MAL067 (Figure 3). Note that "month 3" = 21 days post-499 final dose in MAL068 and MAL071, and 14 days post-final dose in WRAIR 1032; we refer to these 500 slightly disparate post-vaccination time points as "month 3" for simplicity. BTMs with significantly 501 We next performed the baseline correlates analysis of MAL067 (left-most column, Figure 6B). 519 Compared to the results from the month 3 analysis, the baseline correlates analysis of MAL067 520 revealed a larger number of BTMs whose month 0 levels (in vehicle-stimulated PBMC) associated 521 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted May 19, 2021. ; https://doi.org/10.1101/2021.05.19.21257227 doi: medRxiv preprint (nearly all positively) with clinical malaria risk ( Figure 6B, Table S6B). These BTMs spanned many 522 functional categories and were related to DCs, inflammation, and monocytes, among other areas. By far 523 the strongest correlation with risk was with "enriched in monocytes (II) (M11.0)" (FDR = 4.74E-9), 524 followed by "inflammatory response (M33)" (FDR = 1.48E-5) and "resting dendritic cell surface 525 signature (S10)" (FDR = 2.69E-4). Comparing across studies, the greatest similarity seemed to be 526 between MAL067 and WRAIR 1032, which shared 26 common BTMs associated with risk; of these, 527 12 were also associated with risk in MAL068 RRR. Of note, the two functional annotations whose 528 BTMs most consistently associated with risk across these three studies were "Dendritic cells" ("resting 529 dendritic cell surface signature (S10)", "complement and other receptors in DCs (M40)", "DC surface 530 signature (S5)", and "enriched in dendritic cells (M168)" correlated with risk in all three studies) and 531 "Monocytes" ("enriched in monocytes (I)", "enriched in monocytes (II)", "enriched in monocytes 532 "IV", and "monocyte surface signature (S4)" correlated with risk in all three studies). 533 In contrast to the multiple positive baseline transcriptional associations with risk in MAL067, WRAIR 534 1032, and MAL068 RRR, nearly all the month 0 BTMs that associated with challenge outcome in 535 MAL071 RRR were inversely associated with risk, including all the DC-and monocyte-related BTMs 536 seen in the MAL067 signature. We did not find any obvious differences in study population or in 537 sample collection/processing that could explain the differing results. It is also unknown why we see 538 greater consistency in baseline transcriptional associations with risk across studies vs. month 3 539 transcriptional associations with risk. 540 Baseline transcriptional associations with month 3 adaptive responses are presented in Figure S4 and 541 Table S7. No significant associations were seen with any antibody responses. Among other functional 542 BTM categories, baseline levels of three monocyte-related BTMs ("Monocyte surface signature (S4)", 543 "enriched in monocytes (II) (M11.0)", and "enriched in myeloid cells and monocytes (M81)"), as well 544 as two dendritic-cell-related BTMs ("resting dendritic cell surface signature (S10)" and "DC surface 545 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

548
Monocyte frequencies tend to be higher at baseline and at month 3 in RTS,S/AS01 vaccinees that 549 develop clinical malaria 550 The finding that monocyte-related BTMs were expressed significantly higher in RTS,S/AS01 cases vs 551 controls at month 3 in three of the four studies examined ( Figure 6A) and at month 0 in three of the 552 four studies examined ( Figure 6B) suggested that monocyte frequencies may be higher in cases vs 553 controls at these two timepoints. To determine if the monocyte-related transcriptional differences in 554 PBMC were reflected in monocyte populations, the frequencies of monocytes in PBMC cases and 555 controls were compared using immunophenotyping and flow cytometry. Figure S5A shows that, when 556 automated gating (44) is performed, monocyte frequencies at month 0 and, to a lesser extent at month 557 3, tend to be higher in cases than in controls for RTS,S/AS01 and comparator recipients combined. 558 However, these differences were not significant (month 0 p=0.131, month 3 p=0.207). When using 559 manual gating, the inflammatory monocyte frequency and inflammatory monocyte/lymphocyte ratio 560 ( Figure S5B, S5C) also appeared to be higher in cases vs controls, at both month 3 and baseline, but 561 again these differences were not significant. Thus, these findings do not support our hypothesis that the 562 monocyte-related transcriptional differences in case vs. control PBMC would be reflected in these 563 monocyte populations in cases vs. controls. A potential explanation for why we identified a baseline 564 monocyte transcriptional signature of risk yet did not see an association of baseline monocyte 565 frequency, inflammatory frequency, or inflammatory monocyte/lymphocyte ratio with risk, is that the 566 baseline monocyte transcriptional signature of risk reflects expression changes in the existing 567 circulating monocyte population, rather than an expansion in the circulating monocyte population. 568 569 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted May 19, 2021. ;https://doi.org/10.1101https://doi.org/10. /2021 Discussion 570 Our main finding is the identification of a baseline blood transcriptional module (BTM) signature, 571 primarily related to dendritic cells, inflammation, and monocytes, that associates with clinical malaria 572 risk in RTS,S/AS01-vaccinated African children. In a cross-study comparison, much of this baseline 573 risk signature -specifically, dendritic cell-and monocyte-related BTMs -was also recapitulated in two 574 of the three CHMI studies in healthy, malaria-naïve adults. Our finding fits into a growing body of 575 evidence that baseline immune status can influence vaccine responses (45) additional studies of African children in malaria-endemic areas (as our cross-study comparison only 590 looked at CHMI studies in malaria-naïve adults). For instance, an intriguing possibility is that 591 prevaccination administration of anti-inflammatory drugs before vaccine inoculation may potentially 592 reduce baseline inflammation and hence improve RTS,S/AS01-mediated protection. This strategy has 593 been successfully followed in improving antibody responses to the 2012 seasonal influenza vaccine 594 (Agrippal inactivated influenza vaccine) in the elderly (51), albeit it required a 6-week treatment course 595 of the mTOR inhibitor RAD001. While such a treatment course would pose significant logistical 596 challenges in our context, the general approach could potentially become more feasible if new (or 597 existing) anti-inflammatory drugs are found that would require shorter administration periods; there is 598 currently substantial interest towards taking such an approach to improve vaccine response e.g. in older 599 adults (52). 600

601
The role of inflammatory monocytes in vaccine-immunity is an area of active research, however, the 602 association of monocyte activity with clinical malaria risk is consistent with studies that have reported 603 a positive correlation between monocyte to lymphocyte (ML) ratio and clinical malaria risk and/or 604 severity (37, 38). Of note, the association between clinical malaria risk and ML ratio was found to be 605 independent of both age and antibodies to parasite blood-stage antigens ( infection provide a potential basis for this hypothesis. Further studies are needed to determine whether 611 such a inhibitory mechanism is at play in RTS,S vaccine immunity and whether strategies for 612 modulating monocyte populations via chemokine receptor antagonists, as proposed by (55), could help 613 boost RTS,S efficacy. However, here we found that baseline expression of monocyte-related BTMs 614 was positively associated with post-vaccination CSP-specific polyfunctional CD4+ T cell responses. A 615 potential explanation for this association may be more efficient antigen presentation to T cells upon 616 vaccination in individuals with higher baseline expression of these monocyte-related BTMs. The 617 association of month 3 levels of monocyte-related BTMs with risk may be independent from 618 polyfunctional T cells, which we did not find to associate with risk or protection in this work. 619 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted May 19, 2021. ; https://doi.org/10.1101/2021.05.19.21257227 doi: medRxiv preprint 620 Our finding of the association of monocyte signatures with malaria risk seem to contradict the results of 621 a recent study in which we found that monocyte BTMs were associated with protection in RTS,S/AS01 622 vaccinated children and infants (57). However, our previous monocyte-protection association was 623 observed after analyzing gene expression levels in CSP-stimulated, background-corrected (i.e. after 624 subtracting expression in vehicle-stimulated) PBMC (57), whereas the monocyte-risk association 625 described in this study was observed after analyzing gene expression levels in vehicle-stimulated 626 PBMC. Here, we did not detect BTMs associated with the response to RTS,S/AS01 vaccination or with 627 protection when analyzing CSP-stimulated PBMC. This result is not surprising, given the low 628 frequency of CSP-specific T-cells in RTS,S/AS01 vaccinees (e.g. on average, <0.10% of all CD4+ T 629 cells (28)) and the number of T-cell non-responders based on the ICS data reported in this manuscript. 630 In fact, in the previous study we could identify BTMs associated with RTS,S/AS01 vaccination in 631 CSP-stimulated PBMC, but no differentially expressed genes were detected (57) and we had used 632 different stimulation conditions and microarrays instead of RNAseq. 633 634 It is perhaps counterintuitive -considering that the RTS,S/AS01 vaccine does not contain AMA1 -that 635 we observed a small number of BTMs associated with the response to RTS,S/AS01 vaccination and 636 with clinical malaria risk when analyzing AMA1-stimulated PBMC. To explain this result, we refer the 637 reader to our previous work that showed that RTS,S/AS01 vaccination alters antibody responses to 638 antigens not contained in the RTS,S/AS01 vaccine (22). RTS,S/AS01 recipients received partial 639 protection from the RTS,S/AS01 vaccine, leading possibly to decreased P. falciparum parasite load 640 and/or exposure (infection). We hypothesize that the AMA1 stimulation activated T cells that had been 641 previously primed by prior exposure to P. falciparum and that RTS,S/AS01 recipients had fewer 642 primed T cells due to decreased P. falciparum infection (via partial RTS,S/AS01 protection), providing 643 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted May 19, 2021. ;https://doi.org/10.1101https://doi.org/10. /2021 a potential explanation for the transcriptional differences in AMA1-stimulated PBMC between 644 RTS,S/AS01 vs comparator recipients. 645 646 We next discuss some limitations of our study. First, in malaria-naïve adults, the transcriptional 647 response to the third RTS,S/AS01 dose has been shown to peak at Day 1 post-injection, with some 648 decline by Day 6 and approximately 90% of the response having waned by Day 21 (16). Therefore, it is 649 likely that the sampling scheme in this study (one month post-final dose) misses the majority of the 650 transcriptional response to RTS,S/AS01. Future studies with dense, early post-vaccination PBMC 651 sampling could be useful for further investigating RTS,S transcriptional immune correlates. Second, 652 PBMCs were stimulated on site and then frozen. As each site performed the procedure separately, this 653 renders our data susceptible to batch effects. However, a standardized SOP and shared reagents were 654 used, decreasing the possibility of such effects. Moreover, an advantage of on-site stimulation of fresh 655 PBMC is that it avoids the decrease in cell viability, and potential loss of detection of Ag-specific cells, 656 that may have occurred if PBMC had been frozen, thawed, and then stimulated at a central location. 657 Third, there was confounding between age and location. As all infants were from Manhiça and the 658 majority of children were from Bagamoyo, it was not possible to examine the impact of age or clinical 659 trial site on RTS,S/AS01 transcriptional response. Fourth, as only patrolling cell subsets are present in 660 PBMC, we were unable to detect potential signals from T cells, B cells, NK cells, and macrophages 661 localized to an infection site including skin and liver or the immune memory compartment localized in 662 secondary lymphoid organs. 663 664 Despite these limitations, our study also has a number of strengths. For example, while excellent work 665 has already been done to interrogate transcriptional responses to RTS,S/AS01 vaccination in healthy, 666 malaria-naïve adults (including densely sampled early post-vaccination sampling timepoints to capture 667 innate responses) and to identify molecular correlates of RTS,S/AS01-mediated protection against 668 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted May 19, 2021. ;https://doi.org/10.1101https://doi.org/10. /2021 clinical malaria after CHMI in malaria-naïve adults (16,17,43,58), in our study we examined 669 transcriptional responses to RTS,S/AS01 vaccination in infants and children in malaria-endemic areas. 670 This feature is a strength of our study, as 1) infants in particular have relatively immature immune 671 systems (59), making it likely that infants (and younger children) mount different vaccine responses 672 than adults (60); 2) infants and children are especially susceptible to malaria-related morbidity and 673 mortality, making them the target population for this and other malaria vaccines; and 3) continual 674 exposure to P. falciparum, as occurs in endemic areas, influences naturally acquired immunity, which 675 in turn interacts with immunity conferred by RTS,S vaccination (22). Related to this, another advantage 676 of our study is the use of a comparator group which allows to discern the effect of the vaccine from 677 environmental exposures including P. falciparum and age. As participants in the study are very young, 678 significant development of their immune systems occurs throughout the duration of the study, meaning 679 that such changes could potentially be confounded with vaccine-induced immune changes. 680 681 While it will be necessary to perform follow-up studies at more sites and with larger sample sizes to 682 validate the baseline transcriptional signature associated with malaria risk identified here, our study 683 raises interesting hypotheses related to the relationship of inflammation and RTS,S/AS01-mediated 684 protection and suggests potential strategies to explore for augmenting RTS,S/AS01 VE. 685 686 Acknowledgments 687 We are very grateful to study participants, their families, and vaccine trial site field and lab staff. We 688 thank the Phase 3 trial sites PIs Salim Abdulla, Pedro Alonso, Jahit Sacarlal, and Pedro Aide; the 689 investigators involved in the generation of immunology data used here, including providers of antigens 690 for antibody assays (Itziar Ubillos, Marta Vidal, Alfons Jimenez, Ruth Aguilar, Diana Barrios,Laura 691 Puyol, Aintzane Ayestaran, Luis Izquierdo, David Cavanagh, James Beeson, David Lanar, Vir 692 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)  . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)   (no significant  968 correlations were seen with any month 3 antibody responses). Cell color intensity represents the 969 strength of the correlation; BTM/response pairs with significant correlations [false discovery rate 970 (FDR) ≤ 0.2] are outlined in black. Cell color represents correlation direction: red, positive correlation; 971 blue, negative correlation. High-level BTM annotation groups are shown in the left-most color bar. 972 973 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted May 19, 2021. ;https://doi.org/10.1101https://doi.org/10. /2021

986
. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted May 19, 2021. ;https://doi.org/10.1101https://doi.org/10. /2021  . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted May 19, 2021. ;https://doi.org/10.1101https://doi.org/10. /2021 (Tables S3 through S9 are uploaded as separate Excel files) 993 994 Supplementary Table Legends, Tables S3 through S9  995  996  Table S3. List of BTMs, p values, and FDRs for Comparison 1) RTS,S/AS01 vs comparator recipients 997 at month 3 and Comparison 2) RTS,S/AS01 recipients at month 3 vs month 0. 998 Table S4. List of all BTMs tested for significantly different expression in RTS,S/AS01 cases vs 999 controls at month 3, along with stimulation, p value, and FDR results for each Comparison. 1000 Table S5. List of BTMs whose month 3 levels correlated significantly with at least one month 3 1001 adaptive response variable in RTS,S/AS01 vaccinees, along with stimulation, variable details, p value, 1002 and FDR results. 1003 Table S6. List of BTMs whose month 3 or month 0 levels had significantly different expression in 1004 RTS,S/AS01 cases vs. controls in MAL067, along with p values and FDR results when testing the 1005 MAL067 sets of correlate BTMs for significantly different expression in cases vs. controls in the 1006 WRAIR 1032, MAL068 RRR, and MAL071 RRR studies. 1007 Table S7. List of BTMs whose month 0 levels correlated significantly with at least one month 3 1008 adaptive response variable, along with stimulation, variable details, p value, and FDR results. 1009 Table S8. List of all BTMs with significantly different expression in comparator cases vs controls at 1010 month 3, along with p value and FDR results for each Comparison. Only vehicle had any significant 1011 (FDR ≤ 0.2) BTMs. 1012 Table S9. List of BTMs whose month 3 levels correlated significantly with at least one month 3 1013 adaptive response variable in comparator recipients, along with stimulation, variable details, p value, 1014 and FDR results. 1015 1016 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted May 19, 2021. ;https://doi.org/10.1101https://doi.org/10. /2021