Intratumor heterogeneity and clonal distribution of tumor-infiltrating lymphocytes revealed by multiregional T and B cell receptor repertoires in gastric cancer

doi:10.21203/rs.3.rs-2323727/v1

Download PDF

Research Article

Intratumor heterogeneity and clonal distribution of tumor-infiltrating lymphocytes revealed by multiregional T and B cell receptor repertoires in gastric cancer

https://doi.org/10.21203/rs.3.rs-2323727/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Tumor-infiltrating lymphocytes (TILs) recognizing tumor neoantigens can be harnessed as a better prognosis for several human tumors. T and B cell receptor (TCR and BCR) repertoires are the key elements served as tags to track TILs in tumor immune microenvironment. This study intended to simultaneously profile TCR and BCR repertoires via ultra-deep sequencing of multiple tumor regions, adjacent normal mucosa tissues and peripheral blood from five gastric cancer (GC) patients. In parallel, immune-related molecule expression was evaluated for the immunological condition in multiregional tumors by real time-polymerase chain reaction (RT-PCR) and immunohistochemistry. We found that TILs and clonality with majority of T and B cell clones were restricted to each GC regions, obviously different with those of adjacent normal mucosa tissues and peripheral blood, representing T and B cells could react to tumor antigens and delineate the heterogeneity of tumor microenvironment. Especially for BCR repertoires, clonal amplification of B cells dispersedly congested in intratumor regions. There is no correlation between TCR and BCR repertoires in the same GC patient, suggesting that there was a spatio-temporal heterogeneity in GC immune microenvironment. These results suggested that spatially dependent and functionally diverse lymphocyte clones existed in individual GC, expanding our understanding of the adaptive immune response in GC patients.

ultra-deep sequencing

T cell receptor

B cell receptor

intratumor heterogeneity

gastric cancer

Gastric cancer (GC) is one of the most common malignancies worldwide, with a particularly high incidence in Asia [1]. The outcome of most patients with advanced GC remains extremely poor, despite the use of standard therapies such as surgery, radiation, chemotherapy and biological agents [2]. Immunotherapy is a novel treatment for multiple cancer types, including metastatic GC, which involved of adoptive cell therapy (ACT) and immune checkpoint blockade therapy [3–7]. Although the criteria for benefit patients have not been fully understood, immunotherapy still receives widespread attention and investigation. Some studies reported that characteristics of tumor-infiltrating lymphocytes (TILs), tumor mutation burden (TMB) and the expression of PD-1/PD-L1 were associated with immunotherapy response and prognosis [6, 8–10]. Substantial intratumor heterogeneity (ITH) could also be a major challenge for treatment efficacy of immunotherapy [11, 12]. Therefore, systematically and comprehensively recognizing the TILs of ITH is essential for further understanding of the GC immunotherapy.

TILs have been considered as a predictor for tumor progression and surveillance [13, 14]. High-throughput sequencing rendered convenience to analyze the complexities of tumor related immune responses [12, 15–18]. Expressing antigen-specific receptors of B and T lymphocytes would be collected and formed the adaptive immune response system, when exposed to tumor related antigens [19]. TCR and BCR types represent the capacity of T and B cells to critically recognize tumor antigens in tumor microenvironment. Furthermore, the complementary determining region 3 (CDR3) of TCRs and BCRs contributes most to antigen recognition. Large repertoires of T or B cell receptors (TCRs or BCRs, respectively) would be generated by modifying the V regions to enhance the binding affinity towards antigen. Clonal expansion and diversity of T and B cells will play an important role in tumor immunotherapy. As the tumor heterogeneity and intratumor competition of immune responses exist, it is essential to spatially assess the heterogeneous mixture of tumor infiltrating T and B cells [12, 16]. For TCRs, CDR3 mediated TCR interaction with its cognate peptide antigen located within the major histocompatibility complex groove; for BCRs, CDR3 interaction with its cognate polypeptide antigen. Thus, TCR and BCR repertoires analysis could provide a global receptor repertoire of lymphocyte population to a deep extent.

In this study, we used ultra-deep sequencing of the TCR ß and BCR heavy-chain CDR3 in four tumor samples, matched adjacent normal mucosa tissue and peripheral blood from the same five GC patients; we also assessed a panel of immune related genes (Table S1) by quantitative real time-polymerase chain reaction (RT-PCR) in multiple tumor samples, and performed the immunohistochemistry of CD3 and CD19 to evaluate the immune condition in multiple tumor samples. The aims of this study were: (i) to evaluate the correlations of TCR and BCR repertoires in intratumor tissues, adjacent normal mucosa tissues, and peripheral blood in GC patients; (ii) to study whether the intratumoral TCR and BCR repertoires were spatially heterogeneous; (iii) to test the difference and relationship between TCR and BCR repertoire; and (iv) to investigate immune condition from gene level to protein level for our further understanding their diversity and the adaptive responses to tumor antigens in the tumor microenvironment in GC.

Clinical samples

Four different regions of the entire tumor, with its matched adjacent normal mucosa tissues and peripheral blood from each patient, total five GC patients, who were diagnosed with GC and received surgical resection at Peking University Cancer Hospital, Beijing, China. None of these patients had received chemotherapy or radiotherapy before surgery. Each tumor region was cut into three parts, one for DNA extraction, and one for RNA extraction and the rest for immediate formalin fixation.

The detail summary of samples from each patient were listed in Table 1. Peripheral blood was drawn into EDTA-coated tubes prior to administration of anesthesia. DNA was extracted from frozen tissues and blood cells following manufacture’s protocol, using DNeasy Blood & Tissue Kit (Qiagen, Valencia, CA, and USA). The detailed diagram of our experiment was shown in Supplementary Figure S1.

Table 1

Patient clinical characteristics and tissue sample number across different GC patients
Patient	Gender	Age	Tumor size	Tumor location	Differentiation and Lauren	TNM stage	Tumor samples	Matched normal samples	Blood
P1	M	65	1.5*1cm	Stomach angle	Moderate,	T3N0M0	T1, T2, T3, T4	N	B
P1	M	65	1.5*1cm	Stomach angle	adenocarcinoma, Intestinal	T3N0M0	T1, T2, T3, T4	N	B
P2	M	71	10*8cm	Stomach antrum	Poor, adenocarcinoma	T4N3M0	T1, T2, T3, T4	N	B
P2	M	71	10*8cm	Stomach antrum	and partial mucinous, Mix	T4N3M0	T1, T2, T3, T4	N	B
P3	M	53	7.5*7cm	Stomach antrum	Poor,	T4N3M1	T1, T2, T3, T4	N
P3	M	53	7.5*7cm	Stomach antrum	mucinous adenocarcinoma, Mix	T4N3M1	T1, T2, T3, T4	N
P4	M	57	5*4.7cm	Cardiac	Moderate-poor,	T2N0M0	T1, T2, T3, T4	N	B
P4	M	57	5*4.7cm	stomach fundus	adenocarcinoma, Mix	T2N0M0	T1, T2, T3, T4	N	B
P5	F	70	4*3cm	Stomach body	Moderate-poor,	T4N2M0	T1, T2, T3, T4	N	B
P5	F	70	4*3cm	Stomach body	adenocarcinoma, Mix	T4N2M0	T1, T2, T3, T4	N	B
Total							20	5	4

High-throughput Sequencing

The TCR ß and BCR heavy-chain CDR3 regions were amplified using Multiplex PCR in each sample with 1µg per sample amount, and sequenced by Illumina NextSeq 500 (MyGenostics, Beijing, China). The sequencing metrics of TCR and BCR for each sample are summarized in Supplementary Table S2. Sequencing reads were identified according to the definition established by the international ImMunoGeneTics (IMGT) collaboration [20]. Sequencing reads that did not map the structure of the CDR3 region were excluded in the analysis. A standard algorithm was applied to detect V, D and J segments which contributed to the CDR3 region of TCR ß and BCR heavy chain [20]. Those CDR3 sequences containing stop codons or frameshifts with no protein translation were deleted for downstream analyses.

Tcrβ And Bcr Heavy-chain Cdr3 Sequence Analysis

To access the diversity of T and B cell clones in each sample, we utilized the concept of clonality index to depict the distribution of T and B cell clone frequencies [16, 17]. And the Bhattacharyya coefficient to describe the similarity of two TCR and BCR repertoires based on DNA extracted from unfractionated tumor tissues [21]. Clonality was defined as 1-(Shannon’s entropy)/log2 (number of productive unique sequences), which could provide a diversity measurement as a function of clone frequency distribution in each sample and was independent of sampling depth. The maximally diverse population and a perfectly monoclonal population of clonality score would be 0 and 1, respectively [16, 17].

We utilized the Bhattacharyya coefficient to analyze the similarity of two TCR or BCR repertoires, which based on the frequency and homogeneity of shared TCR or BCR sequencing reads within two samples. The formula was as follows:

Bhattacharyya coefficient=\({\sum }_{i=1}^{n}\sqrt{f \left(i, 1\right)\times f\left(i, 2\right),}\)

Where n is the number of overlapping clones and f (i, 1/2) is the frequency of overlapping clones within two TCR or BCR repertoires. 0–1 is the range of the Bhattacharyya coefficient, 0 indicates quite distinct TCR or BCR repertoire and 1 indicates the identical ones between the two samples.

The sequences with stop codons or frameshifts in the CDR3 region were deleted during analysis of the sequencing results. Because of the depth of sequence in the present study, the low-frequency T or B cell clones might be failure to detect. For the analysis of heterogeneity of the TCR or BCR repertoire, we precluded the low-frequency cell clones and selected top 100 T or B cell clones to list their highest frequency in all samples (see Supplementary Table S3).

Rna Extraction And Quantitative Rt-pcr Analysis

The total RNA was extracted using Trizol reagent (Invitrogen, Carlsbad, CA, USA) and cDNA synthesized by M-MLV reverse transcriptase (Promega, Madison, WI, USA) according to the manufacturer's instructions. Gene expression level was determined on ABI 7500 Fast instrument using Power SYBR Green Master Mix (ThermoFisher Scientific Inc, Waltham, USA). Each sample was tested in triplicate, 2^−∆∆CT method was used to analyze the fold change in the gene expression level normalized by GAPDH. The primer sequences of immune related genes were listed in Supplementary Table S1.

Immunohistochemistry

Tissue sections were baked, dewaxed in xylene, and then rehydrated in graded alcohol concentrations. Hydrogen peroxide (3%) was used to block the endogenous peroxidase activity for 10 min. After antigen retrieval and blocking, the sections were incubated with primary antibody overnight at 4℃. The next day, the ready-to-use EnVision reagent (EnVision detection system peroxidase/DAB, rabbit/mouse; Dako, Agilent Technologies, USA) was utilized to bind with the primary antibody. 3.3’-Diaminobenzidine tetra hydrochloride (Dako, Agilent Technologies, USA) to visualize the reaction. As negative controls, the sections were processed using the same protocol except without the primary antibody. The antibodies in this study were: CD3 (working solution, ZM-0417, ZSGB-BIO, CN) and CD19 (working solution, ZM-0038, ZSGB-BIO, CN).

For quantity of CD3 and CD19 staining in tissue sections. Images of stained tissue sections were scanned with Aperio ScanScope imaging system (Leica Aperio LV1, GER) and analyzed with HistoQuest software v3.5.3.0185 (TissueGnostics, AUT). Select five points in each tissue section to analyze CD3 and CD19 staining (one central point with four points around it: up, down, left and right, 0.5×0.5 mm²). The proportion of CD3 and CD19 positive cells were analyzed in tumor and its matched adjacent normal tissues.

Statistical analysis

The SPSS 20.0 software (SPSS, Chicago, USA) was used for statistical analysis. Nonparametric t tests (Mann-Whitney test) or One-way ANOVA (and nonparametric) (Kruskal-Wallis test) were utilized depending on the category of data. We utilized the Spearman’s rank test to evaluate the association between TCR and BCR clonality in each region of all tumor samples. A 2-sided P < 0.05 was considered statistically significant.

High-throughput sequencing of T and B cells

We amplified TCR ß and BCR heavy-chain CDR3 regions by Multiplex PCR, then examined the sequences of each regions using high-throughput sequencing analysis. As shown in Table 2, the average of sequence TCR ß reads was 3.26 million, distributed among a median of 17607 (interquartile range: 13478–30644) unique TCR ß CDR3 rearrangements in all tumor samples. The average productive TCR ß reads in tumor tissue was not significantly different compared with other tissue types (T vs N: P = 0.691; T vs B: P = 0.064, Mann-Whitney test) and a similar result was obtained from average unique TCR ß reads (T vs N: P = 1.000; T vs B: P = 0.286, Mann-Whitney test). Top 100 TCR ß CDR3 sequences from each sample were further assessed by comparing the percentage of repertoire. The average fraction of the top 100 TCR ß CDR3 sequences was 47.6% in tumor tissues, 43.7% in their matched normal tissues and 31.2% in their matched peripheral blood, respectively. And there was no significant difference (P = 0.253 using Kruskal-Wallis test).

Table 2

TCR and BCR CDR3 sequencing metrics in five GC patients
Patient	Region	Clonality	Productive TCR ß reads	Unique TCR ß reads	Highest frequency clone (%)	Top 100 clones (%)
P1	B	0.306682953	2873605	10690	5.02	47.63
	T1	0.318621565	2975487	17148	5.95	46.05
	T2	0.277074092	1686704	12655	4.82	43.09
	T3	0.331270839	2583666	13609	5.85	51.46
	T4	0.31915982	2343483	18391	5.44	46.65
	N	0.313319992	2621128	12317	3.49	48.13
P2	B	0.225077672	3947333	54664	1.35	23.72
	T1	0.258807359	3248839	33020	1.76	30.07
	T2	0.27047756	3540427	29558	2.80	32.95
	T3	0.283343851	2893859	17607	2.49	38.61
	T4	0.32512473	4808645	18135	1.92	45.29
	N	0.311845612	3059351	14895	4.35	47.05
P3	T1	0.460478484	2026498	6617	20.67	79.00
	T2	0.329318141	3032914	21682	5.14	49.21
	T3	0.317321789	3814702	22723	3.07	45.40
	T4	0.383302125	806221	3022	7.49	80.35
	N	0.261531415	2400251	14949	3.70	37.22
P4	B	0.197235537	3849653	58341	0.50	11.58
	T1	0.304492724	2450193	13527	4.97	45.99
	T2	0.293811127	1416545	15035	3.96	43.54
	T3	0.398278311	1867464	3875	15.32	76.27
	T4	0.316631644	3894661	13779	5.74	49.57
	N	0.286523297	2673599	25886	5.87	38.18
P5	B	0.339112159	6491942	38876	9.10	41.90
	T1	0.283749149	3432340	38882	2.36	36.75
	T2	0.289823669	6510283	45779	8.14	34.78
	T3	0.279361969	3057620	31729	5.52	36.03
	T4	0.337602933	1467614	13429	14.91	49.88
	N	0.361151418	8900679	27820	11.60	48.26

For the BCR heavy-chain CDR3 region sequencing data, a median of productive BCR sequences was 4803898 (interquartile range: 3946845–7343789), distributed among a median of 30801 (interquartile range: 20428–39849) unique BCR clones for each sample from five patients. The average of productive BCR reads in tumor tissues was not statistically different compared with other tissue types (T vs. N: P = 0.691; T vs. B: P = 1.000, Mann-Whitney test), and the average unique BCR reads among different samples was no significantly different (P = 0.421 using Kruskal-Wallis test). The average fraction of the top 100 BCR heavy-chain CDR3 sequences was 40.5% in tumor tissues, 26.1% in their matched normal tissues and 61.6% in their matched peripheral blood, respectively. And there was a slightly significant difference (P = 0.069 using Kruskal-Wallis test).

To reduce the sampling bias of rare T and B cell clones, we calculated our downstream analysis only with the top 100 TCR ß and BCR heavy-chain CDR3 sequences in each sample (tumor and matched normal tissues, peripheral blood samples).

Clonal Distribution Of T And B Cells Within Tumor, Matched Normal Tissues And Peripheral Blood

As tumor-infiltrated T or B cells could be expanded towards tumor antigen, the analysis of T or B clonal expansion might demonstrate the heterogeneity of TCR or BCR repertoire and give rise to high frequency of T or B cell clones. We investigated the difference of TCR and BCR clonal distribution across different tissues and multiple regions of each GC samples. Firstly, a metric of clonality was used to measure the clonal distribution: value 0 indicates an even distribution of T or B cell clone frequencies; value 1 represents an asymmetric distribution of T or B cell clone frequencies which indicate some of clones with high frequencies. For T cell clones, the clonality of T cells in tumor microenvironment was slightly higher than that in match normal tissues or peripheral blood with no obvious significant (P = 0.413 for T vs. B; P = 0.841 for T vs. N; Figure. 1A). Furthermore, we continued to compare the TCR clonality across multiple regions of each GC samples. The average clonality of tumor-infiltrating T cells ranged from 0.28 (patient #2) to 0.35 (patient #3) (Figure. 1B). Analysis of the clonality difference of tumor-infiltrating T cells displayed that clonality on percent ranged from 5.42% (patient #1) to 14.32% (patient #3) (Figure. 1C). We also found TCR clonalities in different tissue compartments and multiple regions of the same tumor varied (Figure. 1D). These analysis results revealed ITH of clonality was existed and varied with the different patient individual.

For B cell clones, the clonality of B cells in normal microenvironment was slightly higher than that in match tumor tissues or peripheral blood with no obvious significant (P = 0.191 for T vs. B; P = 0.310 for T vs. N; Figure. 1E). Although the clonality of B cells was not more profuse in tumor tissues compared with normal tissues, we still analyzed the distribution of BCR clonality across multiple regions of each GC samples. The average BCR clonality ranged from 0.30 (patient #5) to 0.41 (patient #3) (Figure. 1F); the percent of clonality difference of BCR ranged from 10.05% (patient #1) to 19.87% (patient #4) (Figure. 1G); we also found BCR clonalities in different tissue compartments and multiple regions of the same tumor varied (Figure. 1H). These analysis results revealed ITH of clonality existed and varied with the different patient individually, and these differences of B cell clones might be originated from disorganized influx from peripheral blood.

Tcr And Bcr Repertoires Analyzed By Bhattacharyya Coefficient Across Different Patients With Gc

We utilized the Bhattacharyya coefficient, which covered both the number and the abundance of overlapping TCR and BCR clones, to investigate the heterogeneity of TCR and BCR repertoires via multiple samples and different sites within the same tumor. The Bhattacharyya coefficient is from 0 to 1: 0 represents completely distinct TCR and BCR repertoires, 1 represents totally identical TCR and BCR repertoires. Firstly, we analyzed pairwise TCR and BCR overlap for each patient and the correlation of TCR and BCR (Figure. 2A and B). Although the intratumoral TCR repertoire was more similar among multiple regions (Bhattacharyya coefficient was from 0.182 to 0.581) within the same tumor compared to tumor versus blood (from < 0.001 to 0.249) or matched normal tissue (from 0.090 to 0.350). The average Bhattacharyya coefficient distributed in multiple tumor regions of each patient was 0.469, 0.374, 0.343, 0.277 and 0.459, respectively. It indicated that the identical TCRs between different tumor samples were also quite lower compared with the heterogeneous TCRs.

However, for the tumor-infiltrating B cells which might originate from an influx of peripheral B cells, we found that the intratumoral B cell repertoire was disorganized among different samples and regions of the same tumor (Figure. 2A). We also analyzed the correlation between TCR and BCR clonality in the five GC patients, unfortunately, no significant correlation was found (P = 0.444; Figure. 2B). These results indicated that T cell clones within GC were still spatial heterogeneity in a single tumor, though different tumor regions shared more TCR ß sequences with each other than with other samples of the patient. All these were not for tumor-infiltrating B cell clones which was inclined to randomly influx from peripheral blood and execute clonal expansion when reacting to tumor antigen.

Furthermore, we compared the mean TCR and BCR repertoire overlaps within tumor tissues (T/T), normal tissues vs tumor tissues (N/T) and peripheral blood vs tumor tissues (B/T) (Figure. 2C and D). For TCR repertoires by comparison between two groups (T/T, N/T, B/T), the average of TCR Bhattacharyya coefficients in “T/T” was significantly higher than those between “N/T”, and “B/T” (P < 0.0001 and < 0.0001, respectively), even that in “N/T” compared with “B/T” (P = 0.010). For BCR repertoires, although the average of BCR Bhattacharyya coefficients was significantly higher in “N/T” and “T/T” when compared with that in “B/T” (P < 0.0001 and < 0.0001, respectively), there was no statistical difference between “N/T” and “T/T” (P = 0.404).

Intratumor Heterogeneity Of Tcr And Bcr Observed Across Multiple Regions Of The Same Tumor

To evaluate the spatial heterogeneity of T and B cell response in GC, we compared different patient samples by displaying the top 10 regional frequency T and B cell clones with the highest regional frequency across the multiple regions of each patient. As shown in figure. 3A, the top 10 T cell clones were quite irregularly distributed in both tumor samples and matched normal tissues, and these clones were seldom detected in their matched peripheral blood. However, most of the top 10 clones were still distributed in all four regions from the same tumor, seldom detected in matched normal tissue of patient #1, and peripheral blood of patient #4.

For BCR repertoire shown in figure. 3B, the majority of top 10 clones of B cell were detected in most of tumor tissues of each patient, but seldom in their matched adjacent normal tissues except patient #3 and #4, and undetected in peripheral blood. The highest top 10 clones were mainly detected in T4 of patient #4, and there still were some top 10 clones distributed in matched normal tissues in patient #5. However, the frequency of the top 10 clones in different tumor regions of the same tumor was quite diverse. These data suggested that high-frequency of tumor-infiltrating T and B cell clones could be detected in tumor tissues, they might be stimulated and expanded when encountering tumor antigens in the tumor microenvironment. Although tumor infiltrating B cells might be randomly perfused with peripheral circulation, it could be produced the more expanded B cell clonotypes towards tumor antigens which limited to the tumor microenvironment, when compared with matched normal tissues. There were no circulated high B cell clones compared with circulated high T cell clones.

Quantitative Rt-pcr And Immunohistochemistry Analysis Revealed Spatial Heterogeneity Of Tissue-related Immunity

We then checked immune-related gene expression level (a 19-gene panel) using quantitative RT-PCR within multiple tumor tissues of these five GC patients (Figure. 4A). 19 immune-related genes were differentially expressed within 4 tumor tissues from one patient. However, the majority of these genes were aggregately detected in the same region of tumor tissues in patient #2 and #3; Patient #1, #4 and #5 displayed an intensive heterogeneity within the intratumor tissues.

Figure 4B-C represented immunohistochemical staining of CD3 and CD19 images of multiple tumor and matched normal tissues. These antibodies were mainly localized on the plasma membrane of T and B cells in these tissues. The tumor-infiltrating T cells of these tissues were roughly stained by CD3; B cells were roughly stained by CD19. CD3⁺ T and CD19⁺ B cells were scattered or clustered across the tumor tissues. For CD19⁺ B cells, we also found that they were mainly distributed near the junction of the tumor and adjacent normal tissues. Furthermore, CD3 and CD19 expression were quantified by TissueFAX cytometry (Figure. 4C). It demonstrated that the expression of CD3 and CD19 was heterogeneously distributed among different samples.

TILs are emerging as a vital biomarker in predicting the efficacy and outcome of immunotherapy treatment, which consists of various lymphocytic cell populations invading the tumor tissues [8, 10, 12]. In our present study, we first showed that the clonality and abundance of intratumoral T and B cell populations were more heterogeneously distributed in intratumor tissues when compared with matched normal tissues and peripheral blood from each of five GC patients. The majority of top 10 T and B cell clones were differentially detected in multiple tumor samples of each patient, but seldom in matched normal tissues and peripheral blood. High-frequency of tumor-infiltrating T and B cell clonalities might be derived from the tumor antigens. Intratumoral TCRβ and BCR heavy-chain CDR3 repertoires were spatially heterogeneous and TIL polyclones were predominantly stimulated by tumor antigens, which were distinctly different from lymphocyte cells circulating in the peripheral blood.

Complicated gene rearrangements and alternative RNA splicing events created cancer cell pathological diversity [22]. The corresponding diverse consequence of the lymphocyte antigen receptor repertoire is estimated more than 10¹² different TCR or BCR proteins, followed the rule of “one cell clone - one receptor sequence” [23, 24]. Rarely two cells came from different lymphocytes with the same antigen–receptor sequence or ‘clonotype’. Thus, we investigated the clonotypes of these tumor infiltrating T and B cells. Previous studies were mainly focused on TCR repertoire, but not on BCR repertoire. As T cells were the direct performer to tumor cells expressed tumor antigens in the cellular immunity, B cells might act as a core player to secret cytokines or directly regulate CD3⁺ T cells in the whole immune network [25–28]. In our study, TCR and BCR repertoires were simultaneously investigated in five GC patients. We found there were more overlaps of TCR or BCR repertoire in intratumor comparisons than that in tumor vs. blood, although there was no directly correlation between TCR and BCR in the same patient. This phenomenon hinted us that we might further explore their interaction with the subgroup of T and B cells. From the demographic high-frequency of tumor-infiltrating B cell clonalities, there were much more oligoclonality of BCR than TCR in the different tumor region, and no high clone of B cells to circulate into peripheral blood compared with high T cell clones. This phenomenon tested high clonal B cells could mainly target on tumor antigens and restricted in the tumor microenvironment. Our approach to understand the involvement of B cells in GC would be useful for uncovering the role of B cells in tumor progression. Previous studies indicated B cells might play an important insight in patients responding to immune checkpoint blockade via targeted expression profiling or bulk RNA sequencing in melanoma and renal cell carcinoma [29, 30]. These studies suggesting antigen-specific humoral responses based on B cell compartment should need us to further investigation in the area of immunotherapy treatment.

Further, individually mining the TCR and BCR repertoires, the top clone of tumor infiltrating T and B cells were both significantly heterogeneous among different tumor regions in one patient, especially the high-frequency of T and B cell clones mainly aggregated in the intratumor tissues. The quantitative RT-PCR results showed some immune-related genes, most of which reflected anti-tumor immunity and correlated with clinical outcome in GC, were also spatially heterogeneous in different tumor regions from the same patient [31–33]. IHC results showed that CD3⁺ T and CD19⁺ B cells were unevenly distributed throughout the tumor tissue, accompanied by aggregating in the mesenchyme of tumor tissues and matched normal tissues. Especially for CD19⁺ B cells, it tended to aggregate in the matched normal tissues or the tumor marginal zone, which pattern of distribution of TILs was the same with that in liver tumor and pancreatic cancer etc. as previously reported [17, 34]. It might be that antigen-stimulated B cells migrate up a chemokine gradient toward the B/T boundary during immune surveillance and in the immune response [35].

Although most of our observations were restricted on the small number of sample sizes, a considerable number of interesting and important findings has been gotten from our data. There are still more fields needed us to explore, such as we just selected TCR ß chain and BCR heavy chain to study on behalf of all characteristics of TCR and BCR, neglecting TCR α chain and BCR light chain. Different types of TILs still need us to comprehensively study and evaluated their roles in prediction of response to immunotherapy [36, 37]. Single-cell RNA-sequencing (scRNA-seq) might be a powerful tool to delineate the single cell distribution and interactions in the tumor microenvironment [38, 39]. However, our results hinted that multiple biopsies might should be analyzed to illustrate the heterogeneity of tumor tissues even for scRNA-seq, and subdivided group of T and B cells needed to study for their further interaction. For the high-clonality of anti-tumor T and B cells, we could utilize their corresponding BCRs or TCRs as an anti-tumor tools like chimeric antigen receptor-modified T (CAR-T) and TCR-modified T (TCR-T) technique to attack tumors.

In conclusion, we first investigated the repertoires of BCR and TCR with high-throughput sequencing to evaluate the clonal composition of TILs across different GC patient samples (tumor, matched normal tissues and blood), detecting from unfractionated sample DNA without prior separation of the T cell population. Our results showed that both compositions of tumor-infiltrating B and T cells in GC were quite different from those of adjacent normal tissues and peripheral blood. High heterogeneity of TCR and BCR repertoires indicated that a high-clonality group of T and B cells were spatially confined to the tumor microenvironment and mainly reacted to tumor antigens, which could give us hints of adaptive immune therapy to deal with GC. Therefore, multiple biopsies might be useful to give us an insight into the adaptive immune response of individual GC patients, and further studies should focus on the subgroup of T and B cells to explore their interaction.

Author Contributions: J.J. performed the study concepts. X.X., X.C., J.H., T.G. and C.Z. performed most of experiments and data analysis and interpretation. Y.H. performed quality control of data and algorithms. H.D., X.L. performed experiments and data acquisition. X.C., J.H., X.X and C.Z. wrote and edited the manuscript.

Conflict of interests: The authors declare no conflict of interests.

Data Availability Statement: The data that support the findings of this study are available from the corresponding authors upon request.

Ethics Statement: This study was approved by the Clinical Research Ethics Committee of Peking University Cancer Hospital & Institute and carried out in accordance with the principles outlined in the Declaration of Helsinki. All patients provided informed consent for obtaining the tissue specimens.

Funding: The National Natural Science Foundation of China (Grant No: 81802471, 81872502 and 81972758); the Science Foundation of Peking University Cancer Hospital (2021-8, 2020-6 and 2020-22); the Third Round of Public Welfare Development and Reform Pilot Projects of Beijing Municipal Medical Research Institute (Beijing Medical Research Institute, 2019-1); Key Program of Joint Funds by National Natural Science Foundation (U20a20371); Beijing Hospitals Authority Youth Program (Grant No: QML 20191105).

Bray F, Ferlay J, Soerjomataram I, et al. Global cancer statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin 2018; 68: 394-424.
Miller KD, Siegel RL, Lin CC, et al. Cancer treatment and survivorship statistics, 2016. CA Cancer J Clin 2016; 66: 271-289.
Bilgin B, Sendur MA, Bulent Akinci M, et al. Targeting the PD-1 pathway: a new hope for gastrointestinal cancers. Curr Med Res Opin 2017; 33: 749-759.
Ott PA, Hodi FS, Robert C. CTLA-4 and PD-1/PD-L1 blockade: new immunotherapeutic modalities with durable clinical benefit in melanoma patients. Clin Cancer Res 2013; 19: 5300-5309.
Restifo NP, Dudley ME, Rosenberg SA. Adoptive immunotherapy for cancer: harnessing the T cell response. Nat Rev Immunol 2012; 12: 269-281.
Kim ST, Cristescu R, Bass AJ, et al. Comprehensive molecular characterization of clinical responses to PD-1 inhibition in metastatic gastric cancer. Nat Med 2018; 24: 1449-1458.
7. Swaika A, Hammond WA, Joseph RW. Current state of anti-PD-L1 and anti-PD-1 agents in cancer therapy. Mol Immunol 2015; 67: 4-17.
Lee HE, Chae SW, Lee YJ, et al. Prognostic implications of type and density of tumour-infiltrating lymphocytes in gastric cancer. Br J Cancer 2008; 99: 1704-1711.
Xing XF, Guo JP, Ding GY, et al. Analysis of PD1, PDL1, PDL2 expression and T cells infiltration in 1014 gastric cancer patients. Oncoimmunology 2018; 7: e1356144.
Samstein RM, Lee CH, Shoushtari AN, et al. Tumor mutational load predicts survival after immunotherapy across multiple cancer types. Nat Genet 2019; 51: 202-206.
Rasmussen JH, Lelkaitis G, Hakansson K, et al. Intratumor heterogeneity of PD-L1 expression in head and neck squamous cell carcinoma. Br J Cancer 2019; 120: 1003-1006.
Reuben A, Gittelman R, Gao JJ, et al. TCR Repertoire Intratumor Heterogeneity in Localized Lung Adenocarcinomas: An Association with Predicted Neoantigen Heterogeneity and Postsurgical Recurrence. Cancer Discov 2017; 7: 1088-1097.
Halpern AC & Schuchter LM. Prognostic models in melanoma. Semin Oncol 1997; 24: S2-7.
Fridman WH, Pages F, Sautes-Fridman C, et al. The immune contexture in human tumours: impact on clinical outcome. Nat Rev Cancer 2012; 12: 298-306.
Gerlinger M, Quezada SA, Peggs KS, et al. Ultra-deep T cell receptor sequencing reveals the complexity and intratumour heterogeneity of T cell clones in renal cell carcinomas. J Pathol 2013; 231: 424-432.
Zhang CT, Huang HY, Miao Y, et al. Clonal distribution and intratumour heterogeneity of the B-cell repertoire in oesophageal squamous cell carcinoma. J Pathol 2018; 246: 323-330.
Cui C, Tian X, Wu J, et al. T cell receptor beta-chain repertoire analysis of tumor-infiltrating lymphocytes in pancreatic cancer. Cancer Sci 2019; 110: 61-71.
Tumeh PC, Harview CL, Yearley JH, et al. PD-1 blockade induces responses by inhibiting adaptive immune resistance. Nature 2014; 515: 568-571.
Ada G. The enunciation and impact of Macfarlane Burnet's clonal selection theory of acquired immunity. Immunol Cell Biol 2008; 86: 116-118.
Yousfi Monod M, Giudicelli V, Chaume D, et al. IMGT/JunctionAnalysis: the first tool for the analysis of the immunoglobulin and T cell receptor complex V-J and V-D-J JUNCTIONs. Bioinformatics 2004; 20: Suppl 1: i379-85.
Rempala GA & Seweryn M. Methods for diversity and overlap analysis in T-cell receptor populations. J Math Biol 2013; 67: 1339-1368.
Willis NA, Rass E, Scully R. Deciphering the Code of the Cancer Genome: Mechanisms of Chromosome Rearrangement. Trends Cancer 2015; 1: 217-230.
Calis JJ & Rosenberg BR. Characterizing immune repertoires by high throughput sequencing: strategies and applications. Trends Immunol 2014; 35: 581-590.
Laydon DJ, Bangham CR, Asquith B. Estimating T-cell repertoire diversity: limitations of classical estimators and a new approach. Philos Trans R Soc Lond B Biol Sci 2015; 370: 20140291.
Hanahan D, Weinberg RA. Hallmarks of cancer: the next generation. Cell 2011; 144: 646-674.
Pardoll DM. The blockade of immune checkpoints in cancer immunotherapy. Nat Rev Cancer 2012; 12: 252-264.
Lund FE., Randall TD. Effector and regulatory B cells: modulators of CD4+ T cell immunity. Nat Rev Immunol 2010; 10: 236-247.
DiLillo DJ, Yanaba K, Tedder TF. B cells are required for optimal CD4+ and CD8+ T cell tumor immunity: therapeutic B cell depletion enhances B16 melanoma growth in mice. J Immunol 2010; 184: 4006-4016.
Helmink BA, Reddy SM, Gao JJ, et al. B cells and tertiary lymphoid structures promote immunotherapy response. Nature 2020; 577: 549-555.
Sautes-Fridman C, Petitprez F, Calderaro J, et al. Tertiary lymphoid structures in the era of cancer immunotherapy. Nat Rev Cancer 2019; 19: 307-325.
Cai XY, Wang XF, Li J, et al. Overexpression of CD39 and high tumoral CD39(+)/CD8(+) ratio are associated with adverse prognosis in resectable gastric cancer. Int J Clin Exp Pathol 2015; 8: 14757-14764.
Ponzoni M, Ferreri AJ, Pruneri G, et al. Prognostic value of bcl-6, CD10 and CD38 immunoreactivity in stage I-II gastric lymphomas: identification of a subset of CD10+ large B-cell lymphomas with a favorable outcome. Int J Cancer 2003; 106: 288-291.
Ito S, Fukagawa T, Noda M, et al. Prognostic Impact of Immune-Related Gene Expression in Preoperative Peripheral Blood from Gastric Cancer Patients. Ann Surg Oncol 2018; 25: 3755-3763.
Giusca SE, Wierzbicki PM, Amalinei C, et al. Comparative analysis of CD4 and CD8 lymphocytes - evidences for different distribution in primary and secondary liver tumors. Folia Histochem Cytobiol 2015; 53: 272-281.
Okada T, Miller MJ, Parker I, et al. Antigen-engaged B cells undergo chemotaxis toward the T zone and form motile conjugates with helper T cells. PLoS Biol 2005; 3: e150.
Ascierto PA, Kalos M, Schaer DA, et al. Biomarkers for immunostimulatory monoclonal antibodies in combination strategies for melanoma and other tumor types. Clin Cancer Res 2013; 19: 1009-1020.
Govers C, Sebestyen Z, Coccoris M, et al. T cell receptor gene therapy: strategies for optimizing transgenic TCR pairing. Trends Mol Med 2010; 16: 77-87.
Zhang M, Hu S, Min M, et al. Dissecting transcriptional heterogeneity in primary gastric adenocarcinoma by single cell RNA sequencing. Gut 2021; 70: 464-475.
Zhang L, Yu X, Zheng L, et al. Lineage tracking reveals dynamic relationships of T cells in colorectal cancer. Nature 2018; 564: 268-272.

Supplementary Material is not available with this version.

No competing interests reported.

Download PDF

Version 1

posted

You are reading this latest preprint version

Intratumor heterogeneity and clonal distribution of tumor-infiltrating lymphocytes revealed by multiregional T and B cell receptor repertoires in gastric cancer

Status:

Version 1

Abstract

Figures

Introduction

Materials And Methods

Clinical samples

High-throughput Sequencing

Tcrβ And Bcr Heavy-chain Cdr3 Sequence Analysis

Rna Extraction And Quantitative Rt-pcr Analysis

Immunohistochemistry

Statistical analysis

Results

High-throughput sequencing of T and B cells

Clonal Distribution Of T And B Cells Within Tumor, Matched Normal Tissues And Peripheral Blood

Tcr And Bcr Repertoires Analyzed By Bhattacharyya Coefficient Across Different Patients With Gc

Intratumor Heterogeneity Of Tcr And Bcr Observed Across Multiple Regions Of The Same Tumor

Quantitative Rt-pcr And Immunohistochemistry Analysis Revealed Spatial Heterogeneity Of Tissue-related Immunity

Discussion

Declarations

References

Supplementary Material

Additional Declarations

Status:

Version 1