How to develop mRNA vaccine? 27+ pancreatic cancer antigen and immune subtype analysis to tell you the answer!

Introduction

A typical cancer vaccine consists of tumor antigens with or without adjuvants , It reprogrammes the immune system to recognize and destroy cancer cells . Immune typing can reflect the comprehensive immune status of tumor and its immune microenvironment , It is closely related to therapeutic response and immune potential .

Background introduction

It is always a hot point to study cancer from the aspect of immunity , This article is brought to you by Xiaobian today , The study found pancreatic cancer （PAAD） Patients have complex tumor immune microenvironment (TIME), by mRNA The development of vaccine and the selection of appropriate vaccinated patients provide a theoretical basis . Published in 《Molecular Cancer》 On , The influence factor is 27.401, The title of the article is ：Identification of tumor antigens and immune subtypes of pancreatic adenocarcinoma for mRNA vaccine development.

Data is introduced

from ICGC and TCGA gain 239 example PAAD Normalized gene expression and clinical follow-up data , as well as 103 Of patients RNA-seq data .

Result analysis

01

PAAD Identification of potential antigens

In order to identify PAAD Potential antigen of , The authors first screened the abnormally expressed genes , And tested 2459 Possible overexpression genes encoding tumor associated antigens ( chart 1a). Then, by analyzing the changed genome fraction and mutation count in a single sample, all the genes that may encode tumor specific antigens are filtered 8850 A mutant gene （ chart 1b and c）. Mutation analysis shows that , In terms of altered genome scores and mutation counts ,KRAS Proto oncogenes GTP enzyme （KRAS） And tumor proteins p53（p53） Is the most common mutant gene （ chart 1d and e）. In actin 、SMAD Family members 4、 High mutation count inhibitors were observed in cyclin dependent kinases, etc 2A（ chart 1e）. A total of 926 Over expressed and frequently mutated tumor specific genes .

chart 1

02

Identification and PAAD Tumor antigens associated with prognosis and antigen presenting cells

Next, the authors screened tumor antigens related to prognosis from the above genes , As a development mRNA Potential candidates for vaccines .21 Genes and PAAD Patient's OS Is closely related to the , among 7 Genes and RFS significant correlation （ chart 2a）. And ADAM9low Group comparison , Overexpression in tumor tissue ADAM Metallopeptidase domain 9（ADAM9） The survival time of patients was significantly shortened （ chart 2b）, Other factors associated with poor prognosis have also been found （ chart 2c-g）.

chart 2

As a result, the right PAAD Development and progress are crucial 6 candidate genes . Besides ,ADAM9、EFNB2、MET and TMOD3 The higher expression level in macrophages 、DC and / or B There was a significant correlation between the increase of tumor infiltration of cells （ chart 3a-d）.TPX2 and WNT7A It also showed an upward trend of increased immune cell infiltration , Although their expression levels vary more （ chart 3e-f）.

chart 3

03

PAAD Identification of potential immune subtypes

Because immunotyping can be used to reflect the immune status of tumor and its microenvironment , To help identify patients who are suitable for vaccination , The author analyzes the data from ICGC Database 239 individual PAAD Sample 1997 Expression profiles of immune related genes , To build consensus clustering . It turns out that immune related genes seem to be stably clustered （ chart 4a and b）, And got it. 5 Immune subtypes , be called IS1-IS5（ chart 4c）.IS1 and IS2 Associated with better prognosis , and IS3 Has the worst probability of survival （ chart 4d）. The distribution of subtypes in different tumor stages and grades shows that , Irregular aggregation of patients diagnosed with different stages （ chart 4e）, and 1 Level and 4 All levels are the same as IS1 significant correlation （ chart 4f）. And ICGC The results obtained by the queue are consistent , The immune subtype is in TCGA There was also a prognostic correlation in the cohort （ chart 4g）, And change significantly at different stages （ chart 4h）, also 1 Level and 4 Both levels show great correlation with IS1 The relevance of （ chart 4i）.

chart 4

04

Association of immune subtypes with mutation status

Due to the high tumor mutation load (TMB) And somatic mutation rate are related to strong anti-cancer immunity , therefore , Author use TCGA Of mutect2-processed Mutation data sets were calculated for each patient TMB And mutation , The same analysis was performed in all immune subtypes . Pictured 5a Shown , And IS1、IS2 and IS3 comparison ,IS4 and IS5 Showing a significantly higher TMB. There is a similar trend in the number of mutant genes （ chart 5b）. Besides , Include KRAS Inside 11 Genes are the most frequently mutated in each subtype （ chart 5c）.

chart 5

05

PAAD Association between immune subtypes and immune modulators

Due to immune checkpoints (ICPs) And immunogenic cell death (ICD) The importance of modulators in cancer immunity , The authors then analyzed their expression levels in different subtypes . Detected in both queues 47 individual ICPs Related genes , among ICGC queue （ chart 6a） Medium 41 individual （87%） Genes and TCGA queue （ chart 6b） Medium 46 individual （97.9%） Genes are differentially expressed among immune subtypes .

It was also found that ,ICGC In line ICP The overall expression level of is higher than TCGA queue . stay ICGC An... Was detected in the queue 28 individual ICD gene , among 22 individual （78.6%） Differentially expressed in immune subtypes （ chart 6c）,25 individual ICD Genes in TCGA Express in a queue , among 24 individual （96%） Showing significant differences between subtypes （ chart 6d）.

chart 6

06

Association between immune subtypes and tumor markers

CA125 and CA199 yes PAAD Established prognostic and diagnostic markers , Higher levels of both indicate cancer progression 、 Poor prognosis or cancer recurrence . In this study ,ICGC and TCGA Cohort in different immune subtypes CA199 and CA125 There are significant differences in expression level . for example , stay ICGC In line ,IS1 as well as IS3 and IS4 Show higher... Respectively CA199 and CA125 expression （ chart 7a and d b）, and IS2 and IS4 stay TCGA Queue with elevated CA199 and CA125（ chart 7c and d）.

chart 7

07

Cellular and molecular characteristics of immune subtypes

Yes mRNA The response of the vaccine depends on the immune status of the tumor . therefore , By using ssGSEA Yes TCGA and ICGC Queue 28 Four previously reported characteristic genes were scored , Further characterizes 5 The immune cell component of the three immune subtypes . Pictured 8a Shown , The immune cell components are divided into five clusters . stay ICGC The cohort showed similar immune cell scores , However, there are significant differences in the composition of immune cells between subtypes （ chart 8b）. therefore ,IS1 and IS2 It's immunity “ heat ” phenotype , and IS4 and IS5 It's immunity “ cold ” phenotype , stay TCGA A similar trend is seen in the queue （ chart 8c and d）.

In order to prove the reliability of this immunotyping , The authors then discussed the five immune subtypes and the previously reported six pan cancer immune subtypes （C1-C6） The correlation between , among PAAD The main aggregation is C1、C2、C3 and C6, Pictured 8e Shown ,IS1 and IS3 Mainly with C3 overlap ,IS2 And C2 and C3 overlap ,IS4 And C1 and C2 overlap ,IS5 And C1、C2 and C6 overlap . Pictured 8f Shown ,IS2 In lymphocytic infiltration 、 White blood cell fraction 、Th1 cells 、 The highest score was in lymphocyte characteristics , But in matrix fraction and TGF-β Low score in response . All in all , Immune subtypes reflect PAAD The cellular and molecular characteristics of the patient , Indicates an immune state .

chart 8

08

PAAD Immune landscape

Patients are used to build PAAD Immune landscape （ chart 9a）. Pictured 9b Shown , Abscissa is related to various immune cells , And the ordinates are mostly immature DC negative correlation . Based on the location of the immune cell population ,IS1、IS4 and IS5 They were further divided into two subgroups （ chart 9c）, The enriched fractions of several immune cells were significantly different among subgroups （ chart 9d）. Besides , Compare the prognosis of samples with extreme distribution in the immune landscape , The first 6 The survival probability of group A was the highest , This is consistent with the above results （ chart 9e and d f）. in summary , The immune landscape based on immune subtypes can accurately identify each PAAD The immune components of patients and predict their prognosis , Conducive to choice mRNA Personalized treatment of vaccines .

chart 9

09

PAAD Immune gene coexpression module and hub Identification of genes

adopt WGCNA（ chart 10a） The samples were clustered to identify immune gene coexpression modules , The soft threshold of scale-free network is 4（ chart 10b and c）. Then the representation matrix is transformed into adjacency matrix , Then it is transformed into a topological matrix . Pictured 10d Shown , As a result 10 Co expression modules , share 1997 A transcript , The genes in the grey module do not cluster with other genes （ chart 10e）. The author further analyzes 9 individual （ Except grey ） The characteristic genes of the module 5 Distribution of immune subtypes , And detected 8 There are significant differences in the distribution of modules （ chart 10f）.

chart 10

Further prognostic correlation analysis showed that , Blue and green modules are associated with PAAD There was a significant correlation with the prognosis of （ chart 11a）. Besides , Blue module and T Cell activation , But with immune landscape components 1 negative correlation （ chart 11b and c）. Again , The green module related to leukocyte migration also showed a consistent negative correlation （ chart 11d and e）. The analysis of prognosis related genes in the blue module shows that , stay ICGC and TCGA In line , Higher expression scores are associated with better prognosis , This is consistent with the above findings （ chart 11f and g）. Last , Three correlations with the blue module have been identified > 90% The central gene of , Include MAP 4 K1、TBC1D10C and TRAF3IP3, They are mRNA Potential biomarkers for vaccines .

chart 11

Editor's summary

The study was conducted in PAAD Six overexpressed and mutated tumor antigens associated with poor prognosis and antigen-presenting cell infiltration were identified , To determine the PAAD Five immune subtypes of (IS1-IS5) And nine immune gene modules , The molecules of immune subtypes were studied 、 Cellular and clinical features , And observed PAAD Immune landscape . This study is a very comprehensive analysis from the perspective of immunity , It's worth learning from ！