Molecular Profile of a GM-CSF Overexpressing Breast Cancer Whole-Cell Vaccine with Systemic Anti-Tumor Activity Markus D

Molecular profile of a GM-CSF overexpressing breast cancer whole-cell vaccine with systemic anti-tumor activity Markus D. Lacher1,*,Charles L. Wiseman1, and Joseph Wagner1 1BriaCell Therapeutics Corp., Berkeley, CA, USA; *presenting author #2369

ABSTRACT PERSPECTIVE RESULTS DISCUSSION

• BriaVax™ is a whole-cell “GVAX” vaccine prepared from a breast cancer cell line with an BriaVax™ (SV-BR-1-GM) HLA-A B2M HLA Types of BriaVax™ and Phase I Clinical Trial Subjects BACKGROUND AND PURPOSE OF STUDY: unusual variety of cytogenetic abnormalities (Wiseman and Kharazi, 2006 and 2010). SUMMARY CP Lot IV add’l expansion CP Lot IV 4p cryo + culture cryo + culture BriaVax™ Other Breast HMECs BriaVax™ Other Breast HMECs Subject Survival Tumor The allogeneic whole-cell cancer vaccine BriaVax™ (formerly SV-BR-1-GM) is an ER/PR +CSF2* • In a small initial clinical trial, one “Special Responder” (SP) experienced prompt, 25000 Cancer Cell Lines 25000 Cancer Cell Lines HLA-A HLA-B HLA-DRB3 The cancer vaccine BriaVax™ expresses Class I and Class II MHC SV-BR-1 SV-BR-1-GM Master Cell Bank (MCB) CP Lot V DCs DCs negative, HER2/neu positive breast cancer cell line (SV-BR-1) we engineered to stably cryo cryo ID (months) regression widespread, and replicable regression at multiple sites of metastatic breast cancer (Wiseman 20000 alleles: CP Lot VIII 20000 overexpress GM-CSF. BriaVax, rendered proliferation incompetent by irradiation, has thus far and Kharazi, 2006). N/A A001 40.7 No 02:01 24:02 13:02 41:01 03:01 - • HLA-A: 11:01 and 24:02 Lot II Lot IV 15000 15000 cryo cryo • HLA-B: 35:08 and 55:01 been applied to 4 advanced stage cancer patients (3 subjects with breast cancer, 1 subject A002 33.7 Yes 02:01 11:01 18:03 44:02 02:02 - • Molecular analysis of HLA, known tumor-associated antigens, and immune response genes 10000 with ovarian cancer). One breast cancer subject responded to BriaVax with complete *stable transfection with 10000 • HLA-DRB3: 01:01 and 02:02 is a necessary effort in addressing the BriaVax™ mechanism of action. We analyzed gene CSF2 (GM-CSF) A003 35.6 No 02:01 03:01 07:02 13:02 Negative - remission of a measurable lung lesion and near complete remission of multiple breast lesions 5000 5000 expression profiles of BriaVax™ using Illumina BeadChip microarray and NanoString Figure 1. BriaVax™ samples used for this study. SV-BR-1-GM (BriaVax™) was derived from SV-BR-1 breast cancer 2 of these alleles were also found in blood cells of the Special

after only 3 inoculations. Nevertheless, she relapsed 3 months after completing the protocol, 0 0 B001 7.0 No 11:01 - 35:01 40:01 Negative -

cryo cryo cryo cryo cryo cryo cryo cryo cryo cryo MCB

nCounter technologies, and compared the data with published results from 16 established cells following stable transfection with CSF2, encoding granulocyte-macrophage colony-stimulating factor (GM-CSF) MCB Responder

Lot II II Lot II Lot Lot IV IV Lot

with brain metastases as well as multiple breast lesions. We obtained FDA permission to (Wiseman and Kharazi, 2006 and 2010). The diagram only shows BriaVax™ samples used for this study. “cryo” refers to IV Lot BriaVax N/A N/A 11:01 24:02 35:08 55:01 01:01 02:02

BriaVax BriaVax CP Lot V V Lot CP

breast cancer cell lines. V Lot CP

CP Lot IV IV Lot CP IV Lot CP

CP Lot IV IV Lot CPculture IV Lot CPculture

CP Lot IV 4p IV Lot CP 4p IV Lot CP T47D_GSE36693

T47D_GSE36693 • HLA-A: 11:01

BriaVax BriaVax

MCF7_GSE36693 MCF7_GSE48398 MCF7_GSE54326 MCF7_GSE36693 MCF7_GSE48398 MCF7_GSE54326

BT474_GSE36693 BT549_GSE36693 BT474_GSE36693 BT549_GSE36693 BriaVax

samples lysed for RNA extraction directly from cryovials, “culture” indicates culturing prior to lysis. BriaVax

MB415_GSE36693 MB468_GSE48398 MB415_GSE36693 MB468_GSE48398

MB157_GSE36693 MB231_GSE36693 MB231_GSE48398 MB231_GSE54326 MB361_GSE36693 MB436_GSE36693 MB468_GSE36693 MB157_GSE36693 MB231_GSE36693 MB231_GSE48398 MB231_GSE54326 MB361_GSE36693 MB436_GSE36693 MB468_GSE36693

SKBR3_GSE54326 SKBR3_GSE54326

ZR75.1_GSE54326 ZR75.1_GSE54326

CP Lot IV 4p culture IV Lot CP 4p culture IV Lot CP

HS578T_GSE36693 HS578T_GSE36693

MCF10A_GSE48398 MCF10A_GSE48398

BriaVax BriaVax

HCC1806_GSE36693 HCC1806_GSE36693 HCC1937_GSE36693 HCC1937_GSE36693

resume vaccinations, and, upon doing so, all metastatic sites responded with a prompt tumor HCC1806_GSE36693

BriaVax BriaVax

BriaVax BriaVax BriaVax

• Histocompatibility allele match(es) between BriaVax™ and patients may be a requirement for BriaVax Table 1. BriaVax™ and Subject A002 with tumor regressions even at metastatic sites (reported in • HLA-DRB3: 02:02

BriaVax BriaVax

senescent HMEC_GSE56718senescent HMEC_GSE56718senescent

DCs_00h_Infected_GSE58278 DCs_06h_Infected_GSE58278 DCs_12h_Infected_GSE58278 DCs_18h_Infected_GSE58278 DCs_24h_Infected_GSE58278 DCs_00h_Infected_GSE58278 DCs_06h_Infected_GSE58278 DCs_12h_Infected_GSE58278 DCs_18h_Infected_GSE58278 DCs_24h_Infected_GSE58278 proliferating HMEC_GSE56718 proliferating

regression after only 3 inoculations. To prospectively identify patients with a high likelihood of HMEC_GSE56718 proliferating DCs_00h_Uninfected_GSE58278 therapy efficacy assuming a mechanism of action in which patient T cells are activated via HLA-A (ILMN_1671054) B2M (ILMN_1725427) DCs_00h_Uninfected_GSE58278 Wiseman and Kharazi, 2006) share both MHC class I (HLA-A) and class II (HLA-DRB3) alleles. benefiting from BriaVax therapy we began a program to identify molecular factors of HLA-A (ILMN_2203950) B2M (ILMN_2148459) cancer antigens co-expressed in BriaVax™ and patient tumors and displayed on BriaVax™ Background (see Fig. Legend) Background (see fig. legend) BriaVax™ overexpresses ERBB2 (Her2/neu) and at least 20 other diagnostic potential. Here, we describe a gene expression signature that might both be Genes up-regulated in SV-BR-1-GM A MHCs. A (compared to normal breast cells) candidate tumor-associated antigens, and expresses several HLA-DM HLM-DRA CD4+ BriaVax™ as APCs informative about BriaVax’ mechanism of action and helpful for developing diagnostic or Th cell Low stringency Medium stringency cancer/testis antigens including PRAME. filter filter Genes/Alleles Expressed in BriaVax™ monitoring biomarkers. BriaVax™ Other Breast MHC II BriaVax™ Other Breast HMECs HMECs HLA-A*11:01 Hypothetical Mechanism of Action BriaVax™ (SV-BR-1-GM) Cancer Cell Lines ? Microarray 995 439 100000 Cancer Cell Lines 100000 IL-6 HLA-A*24:02 Normal breast cells (GSE35399, GEO) DCs DCs CD8+ Several immune response mediators were also identified in the Illumina HumanHT-12 v4 BeadChip genes genes B2M (b2-microglobulin) Normal breast cells (GSE56718, GEO) ? CTL (47323 probes, >31000 annotated genes) 10000 10000 BriaVax™ cell line: CD83 HLA-DRA METHODS: BriaVax™ expresses complete sets of genes predicting presence of both MHC class I (b2- HLA-DR HLA-DRB3*01:01 1000 1000 HLA-DM • CD83: a dendritic cell marker B CD83 To prospectively identify patients with tumors responsive to BriaVax we began a molecular microglobulin, HLA-A, HLA-B) and class II (HLA-DRA, -DRB3, -DMA, -DMB) complexes HLA-DRB3*02:02 Validation (high stringency filter) 100 • IL6: a secreted factor with both pro- and anti-inflammatory properties and presents with high transcript levels of CD83, CD74, and IL6, i.e., factors with 439 100 MHC I analysis of both the BriaVax cell line and cells obtained from the special clinical responder’s genes HLA-DMA • CD74: promotes MHC class II presentation of exogenous antigens via its invariant chain 10 10 CLIP HLA-DMB established roles in T cell activation. Microarray GSE29431 (Breast cancer tissues) (Ii) and CLIP products. Expression in BriaVax™ ~10X higher than in other breast blood. BriaVax gene expression profiles were obtained through Illumina BeadArray and GSE7307 (Nonmalignant tissues from different organs) SV-BR-1-GM Affymetrix Human Genome U133 Plus 2.0 Array 1 Proteases

21 1 (BriaVax™) cancer cell lines and ~10X lower than in DCs. No/low expression possibly better

cryo cryo cryo cryo cryo cryo cryo cryo cryo cryo MCB

NanoString nCounter technologies and compared to gene expression data sets publically MCB Proteasome Lot II II Lot

Strikingly, both BriaVax™ and the special responder carried HLA-A*11:01 and HLA- genes II Lot than high expression for whole-cell vaccines (Thompson et al, 2008). Lot IV IV Lot

Lot IV IV Lot Antigen

BriaVax BriaVax

CP Lot V V Lot CP V Lot CP

CP Lot IV IV Lot CP IV Lot CP CP Lot IV IV Lot CPculture

available through the Gene Expression Omnibus (GEO; National Center for Biotechnology IV Lot CPculture CP Lot IV 4p IV Lot CP

Probe ID Gene Symbol Description 4p IV Lot CP

T47D_GSE36693 T47D_GSE36693

BriaVax BriaVax

MCF7_GSE36693 MCF7_GSE48398 MCF7_GSE54326 MCF7_GSE36693 MCF7_GSE48398 MCF7_GSE54326

BT474_GSE36693 BT549_GSE36693 BT474_GSE36693 BT549_GSE36693 BriaVax

DRB3*02:02 alleles. Furthermore, compared to normal human breast cells, BriaVax™ BriaVax

MB415_GSE36693 MB468_GSE48398 MB415_GSE36693 MB468_GSE48398

SKBR3_GSE54326 SKBR3_GSE54326

ZR75.1_GSE54326 ZR75.1_GSE54326 CP Lot IV 4p culture IV Lot CP

C 4p culture IV Lot CP HS578T_GSE36693

HS578T_GSE36693 Antigen MCF10A_GSE48398

MCF10A_GSE48398 Endosome

BriaVax BriaVax

HCC1937_GSE36693 HCC1937_GSE36693 HCC1806_GSE36693 HCC1937_GSE36693

203545_at ALG8 asparagine-linked glycosylation 8, alpha-1,3-glucosyltransferase homolog (S. cerevisiae), ALG8 HCC1806_GSE36693

BriaVax BriaVax

BriaVax BriaVax BriaVax

Information) portal. overexpresses several genes known to encode TAAs, such as PRAME, a cancer/testis 226914_at ARPC5L actin related protein 2/3 complex, subunit 5-like, ARPC5L BriaVax

BriaVax BriaVax senescent HMEC_GSE56718senescent

senescent HMEC_GSE56718senescent CD8+

230535_s_at ATP5E ATP synthase, H+ transporting, mitochondrial F1 complex, epsilon subunit, ATP5E HMEC_GSE56718 proliferating HLA-DRA (ILMN_1689655) DCs_00h_Uninfected_GSE58278 HLA-DMA (ILMN_1695311) DCs_00h_Uninfected_GSE58278 (pre-primed antigen gene. 208079_s_at AURKA aurora kinase A, AURKA HLA-DMB (ILMN_1761733) HLA-DRA (ILMN_2157441) CONCLUSIONS 204092_s_at AURKA aurora kinase A, AURKA Background (see Fig. Legend) Background (see Fig. Legend) GM-CSF or naïve) RESULTS: 224447_s_at C17orf37 chromosome 17 open reading frame 37, C17orf37 DC activity ↑ Because BriaVax™ 1) expresses genes known for their immune stimulatory roles, and 2), 226473_at CBX2 chromobox homolog 2, CBX2 BriaVax™ expresses a gene signature consistent with a mechanism of action involving not 226237_at COL8A1 collagen, type VIII, alpha 1, COL8A1 While confirmatory studies will be required, the data implies the following: encodes identical MHC I and II alleles as the special responder, we hypothesized that: 230679_at DCAF10 DDB1 and CUL4 associated factor 10, DCAF10 CD74 (invariant chain & CLIP) CD83 degradation only the activation of cytotoxic T cells but also the induction of a humoral response. In 226511_at DCAF10 DDB1 and CUL4 associated factor 10, DCAF10 . BriaVax™ cells can act directly as allogeneic APCs. 234354_x_at ERBB2 v-erb-b2 erythroblastic leukemia viral oncogene homolog 2, neuro/glioblastoma derived oncogene homolog (avian), ERBB2 addition, BriaVax™ expresses known cancer antigens. Notably, blood-derived cells of the 216836_s_at ERBB2 v-erb-b2 erythroblastic leukemia viral oncogene homolog 2, neuro/glioblastoma derived oncogene homolog (avian), ERBB2 BriaVax™ Other Breast HMECs BriaVax™ Other Breast HMECs B 1. The presence of Class I and Class II HLA alleles is consistent with the development . patient DCs can cross-dress with BriaVax™ pMHCs. 1555997_s_at IGFBP5 insulin-like growth factor binding protein 5, IGFBP5 Cancer Cell Lines Cancer Cell Lines 100000 7000 special clinical responder expressed genes complementing BriaVax’ gene expression 218783_at INTS7 integrator complex subunit 7, INTS7 DCs DCs of both cell-mediated and humoral immune responses. . patient DCs can cross-present BriaVax™ antigens on their MHC system. 228491_at KRT19 Keratin 19, KRT19 6000 10000 signature, thus possibly explaining the unusually prompt and robust clinical response to 213711_at KRT81 keratin 81, KRT81 5000 2. The double match of HLA alleles between the Special Responder and BriaVax™ is 202903_at LSM5 LSM5 homolog, U6 small nuclear RNA associated (S. cerevisiae), LSM5 BriaVax™. 202904_s_at LSM5 LSM5 homolog, U6 small nuclear RNA associated (S. cerevisiae), LSM5 1000 4000 consistent with stringent MHC restriction. 218398_at MRPS30 mitochondrial ribosomal protein S30, MRPS30 3000 ABBREVIATIONS 222824_at NUDT5 nudix (nucleoside diphosphate linked moiety X)-type motif 5, NUDT5 100 221811_at PGAP3 post-GPI attachment to proteins 3, PGAP3 2000 Dendritic Cell 3. The BriaVax™ cancer vaccine may influence the host immune response by: CONCLUSIONS: 55616_at PGAP3 post-GPI attachment to proteins 3, PGAP3 10 Cross-Presentation APC: Antigen-Presenting Cell MCB: Master Cell Bank 1000 (DC) 204517_at PPIC peptidylprolyl isomerase C (cyclophilin C), PPIC • acting as allogeneic antigen-presenting cells (APCs)

Our findings suggest that BriaVax exerts its therapeutic effects via multiple modes. We CP: Cell Product MHC: Major Histocompatibility Complex 205449_at SAC3D1 SAC3 domain containing 1, SAC3D1 1 0

cryo cryo cryo cryo cryo cryo cryo cryo cryo cryo MCB

1557458_s_at SHB Src homology 2 domain containing adaptor protein B, SHB MCB and/or Lot II II Lot

identified both candidate immunogens overexpressed in BriaVax™ compared to normal CTA: Cancer/Testis Antigen NCBI: National Center for Biotechnology Information 225765_at TNPO1 transportin 1, TNPO1 II Lot Lot IV IV Lot

Lot IV IV Lot C Cross-Dressing

BriaVax BriaVax

CP Lot V V Lot CP V Lot CP CP Lot IV IV Lot CP

CP Lot IV IV Lot CP • up-regulating T cell cytotoxicity by cross-presentation and/or “cross-dressing”

CP Lot IV IV Lot CPculture IV Lot CPculture

CP Lot IV 4p IV Lot CP 4p IV Lot CP

T47D_GSE36693 T47D_GSE36693

BriaVax BriaVax

MCF7_GSE48398 MCF7_GSE54326 MCF7_GSE36693 MCF7_GSE48398 MCF7_GSE54326

CTL: Cytotoxic T Lymphocyte pMHC: Peptide-loaded MHC MCF7_GSE36693

BT474_GSE36693 BT549_GSE36693 BT474_GSE36693 BT549_GSE36693

BriaVax BriaVax

MB415_GSE36693 MB468_GSE48398 MB415_GSE36693 MB468_GSE48398

SKBR3_GSE54326 SKBR3_GSE54326

ZR75.1_GSE54326 ZR75.1_GSE54326 CP Lot IV 4p culture IV Lot CP

breast cells and unraveled a potential mechanism of action explaining the encouraging clinical 4p culture IV Lot CP

HS578T_GSE36693 HS578T_GSE36693

MCF10A_GSE48398 MCF10A_GSE48398

BriaVax BriaVax

HCC1806_GSE36693 HCC1806_GSE36693 HCC1937_GSE36693 HCC1806_GSE36693 HCC1937_GSE36693 BriaVax

Figure 2. Identification of candidate immunogens. A. BriaVax™ RNA samples were hybridized onto Illumina HumanHT-12 BriaVax BriaVax

BriaVax and/or BriaVax

DC: Dendritic Cell SP: Special Responder BriaVax

BriaVax BriaVax senescent HMEC_GSE56718senescent

response observed. v4 BeadChip arrays then expression values compared to those of normal human breast cells provided in the Gene Expression HMEC_GSE56718senescent

CD74 (ILMN_1736567) HMEC_GSE56718 proliferating DCs_00h_Uninfected_GSE58278 GEO: Gene Expression Omnibus TAA: Tumor-Associated Antigen DCs_00h_Uninfected_GSE58278 • secreting IL6 and GM-CSF Omnibus (GEO, NCBI) database as DataSets GSE35399 and GSE56718. Two serial filters were applied to the quantile- CD74 (ILMN_2379644) CD83 (ILMN_1780582) Background (see Fig. Legend) HLA: Human Leukocyte Antigen Th cell: T helper cell normalized expression values to enrich for genes likely differentiating BriaVax™ from normal breast cells. After the first, low CD83 (ILMN_2328666) stringency, filter, 995 genes (including 3 unannotated Illumina probes) were retained, of which after the second, medium Background (see Fig. Legend) CD4+ (pre-primed stringency, filter, 439 genes were selected. B. The 439 genes retained after the medium-stringency filter were in silico IL6 or naïve) validated on GEO DataSets GSE29431 (breast cancer tissues) and GSE7307 (nonmalignant tissues representing various HLA-DR (nCounter) organs). This high stringency filtration step served to identify genes with expression levels higher in breast cancer than in a CD8+ (pre-primed METHODS variety of nonmalignant tissues. 21 genes were retained in the high-stringency filter. BriaVax™ Other Breast HMECs HLA-DRA HLA-DRB3 6000 or naïve) REFERENCES Cancer Cell Lines 5000 DCs 300

4000 250 • Lowe et al., Genome Biol. 2015;16:194. Microarray gene expression profiling NanoString nCounter-based gene profiling A BriaVax™ Normal Breast Cells B SPA17 TTK 200 Figure 5. Model of Proposed Mechanism of Action. 3000 • Jongeneel et al., Genome Res. 2005;15(7):1007-14. Expression of the gene signature shown for BriaVax™ and outlined in Figure 4 is HMEC** 120 250 150 • Reich et al., Nat Genet. 2006;38(5):500-1. BriaVax™ cells obtained directly from cryovials following recovery from liquid nitrogen were 2000 D consistent the following mechanisms of anti-tumor immune stimulation: • BriaVax™ cells, obtained directly from cryovials following recovery from liquid nitrogen, or 119 100 Patient tumor • Shehata et al., Breast Cancer Res. 2012;14(5):R134. 200 lysed in Buffer RLT (RNeasy Mini kit, Qiagen) then total RNA isolated from a portion of the LUMINAL* BASAL* STROMAL* 118 1000 A. BriaVax™ as a direct activator of pre-primed T cells. harvested from cultures, were subjected to total RNA extraction (RNeasy Mini kits; 50 Cancer Antigen • Thompson et al., Cancer Immunol Immunother. 2008;57(3):389-98. lysates via RNeasy Mini Kit technology (Qiagen). Samples were hybridized onto the 117 Factors and (some of) their known roles as immune stimulants identified by mRNA

Qiagen, Valencia, CA). Gene expression profiles were established using HumanHT-12 v4 ERBB3+ ERBB3+ ERBB3- Nonclonogenic 150 0 0 • Wiseman and Kharazi, Breast J. 2006;12(5):475-80.

cryo cryo cryo cryo 116 cryo MCB gene expression profiling. Expression of MHC class I and II genes in BriaVax™ is

nCounter® PanCancer Immune Profiling Panel (NanoString Technologies, Seattle, WA). Data ALDH- ALDH+ ALDH- Lot II II Lot

Expression BeadChip arrays (Illumina). Array hybridization and data acquisition was 115 IV Lot consistent with a model in which BriaVax™ directly stimulates CD8+ CTLs and CD4 • Wiseman and Kharazi, The Open Breast Cancer Journal 02/2010; 2(1):4-11. BriaVax

100 V Lot CP

was analyzed using nSolver Analysis Software (NanoString Technologies) and Microsoft IV Lot CP

CP Lot IV IV Lot CPculture

CP Lot IV 4p IV Lot CP

T47D_GSE36693

BriaVax

MCF7_GSE48398 MCF7_GSE54326

114 MCF7_GSE36693

BT474_GSE36693 BT549_GSE36693

BriaVax T helper (T ) cells and thereby, potentially, induces both cytotoxic and humoral

MB415_GSE36693 MB468_GSE48398

MB157_GSE36693 MB231_GSE36693 MB231_GSE48398 MB231_GSE54326 MB361_GSE36693 MB436_GSE36693 MB468_GSE36693

conducted at the University of Minnesota Genomics Center. SKBR3_GSE54326 ZR75.1_GSE54326

CP Lot IV 4p culture IV Lot CP h

HS578T_GSE36693

MCF10A_GSE48398

subtr., HK normalized counts normalizedHK subtr.,

BriaVax

HCC1806_GSE36693 HCC1806_GSE36693 HCC1937_GSE36693

- - Excel. BriaVax

113 BriaVax

50 BriaVax (antibody-based) responses.

cryo

BriaVax

senescent HMEC_GSE56718senescent

DCs_00h_Infected_GSE58278 DCs_06h_Infected_GSE58278 DCs_12h_Infected_GSE58278 DCs_18h_Infected_GSE58278 DCs_24h_Infected_GSE58278

proliferating HMEC_GSE56718 proliferating Bckgr. Bckgr. 112 DCs_00h_Uninfected_GSE58278 B. Cross-presentation of BriaVax™ peptides on DCs

• Non-normalized data was analyzed with various modules of GenePattern using the public 111 0 IL6 (ILMN_1699651) BriaVax CP Lot V Lot CP BriaVax

Background (see Fig. Legend) cryo MCB BriaVax BriaVax™ is degraded and fragments of apoptotic cells taken up by dendritic cells server portal (http://www.broadinstitute.org/cancer/software/genepattern/) (Reich et al., HLA Typing cryo IV Lot BriaVax ACKNOWLEDGEMENTS (DCs) from the patient. BriaVax™ antigens are proteolytically degraded inside such 2006). For some analyses, BriaVax™ samples were compared to other gene expression Figure 4. Immune Stimulators Expressed in BriaVax™ (“Gene Signature”). CEP55 DCs then presented on cell surface MHCs to patient T cells. BriaVax™ and peripheral blood lymphocyte (PBL) samples from 4 subjects of a clinical study PBK data sets also generated via HumanHT-12 v4 BeadChip arrays and available through the PLAC1 • Expression levels of 16 different breast cancer cell lines were compared with those of BriaVax™. Multiple studies GEO (NCBI) portal. Data sets to be compared to one another were merged with the (FDA protocol BB-IND 10312) were subjected to high-resolution HLA typing for HLA-A, HLA-B, PRAME C. Cross-dressing of DCs with BriaVax™ Peptide-MHCs *Shehata et al. (2012). GEO DataSet GSE35399. (different “GSE numbers” from GEO, NCBI) were included for MB231 (MDA-MB-231), MB468 (MDA-MB-468) Allogeneic BriaVax™ cell surface MHCs loaded with BriaVax™ antigens are directly We are grateful to Gerhard Bauer, Brian Fury, Emily Lynn Fledderman, Tye Daniel

and HLA-DRB3 (City of Hope Laboratories, Duarte, CA). **Lowe et al. (2015). GEO DataSet GSE56718. BriaVax MCB cryo MCB BriaVax

MergeColumns module then quantile-normalized using the IlluminaNormalizer module cryo MCB BriaVax and MCF7. HMECs: human mammary epithelial cells (GSE56718), DC: dendritic cells (GSE58278). BriaVax Lot IV cryo IV Lot BriaVax BriaVax Lot IV cryo IV Lot BriaVax transferred onto the cell surface of patient DCs by trogocytosis. T helper (Th) cells Petrie, Dane P. Coleal-Bergum, and Tia M. Hackett from the UC Davis GMP facility

BriaVax CP Lot V cryo V Lot CP BriaVax • Background was defined as the median expression value across all human RNA targeting, non-control, probes (beta version), or, for Affymetrix data sets (in silico validation), using the cryo V Lot CP BriaVax CD8+ (CD4+) recognizing MHC II complexes may induce a humoral anti-tumor response (very rough estimation that approx. 50% of the genes in a tissue are expressed (Jongeneel et al., 2005)). (Sacramento, CA) for their BriaVax™ manufacturing efforts. We thank Darrell Johnson ExpressionFileCreator module. Normalized data was further processed in Microsoft Figure 3. Cancer/Testis Antigens (CTAs) expressed in BriaVax™ A. CTA mRNA expression in BriaVax™ and CTL and CTLs (CD8+) recognizing MHC I complexes a cytotoxic response. • Values on the Y-axes represent normalized gene expression levels. Shown are arithmetic means across and his team from the University of Minnesota Genomics Center for their outstanding normal breast cells (GEO DataSets GSE35399 (23) and GSE56718 (samples GSM1367274, GSM1367276, GSM1367278 Excel. Hierarchical clustering was conducted via the HierarchicalClustering module. Heat replicates per study. D. Patient tumor attack only) as assessed by microarray (Illumina Human HT-12 v4 BeadArray). B. CTA mRNA expression levels (of SPA17 and microarray service. maps of clusters were visualized using the HierarchicalClusteringViewer module and • “HLA-DR (nCounter)” refers to expression levels (counts) as determined via the nCounter (NanoString) Directly (A) or indirectly (C and D) by BriaVax™ activated T cells recognize and kill TTK) as determined via the nCounter platform (NanoString Technologies, Inc.). Visualizer (Gene Pattern). technology. Other panels show expression levels as determined via Illumina microarrays. tumor cells if they express and present cancer antigen(s) that are also expressed in BriaVax™. Additionally, tumor destruction may occur via antibodies.

Printed by