Development And Applications Of Phenotype Defined Spatial Transcriptomic Technology Based On Laser Capture Microdissection With Immunofluorescence

The functioning of tissues is fundamentally dependent upon not only the phenotypes of the constituent cells but also their spatial organization in the tissue,as local interactions precipitate intra-cellular events that often lead to changes in expression.However,our understanding of these processes in native tissues,whether healthy or diseased,is limited at present owing to the difficulties in acquiring comprehensive transcriptional programs with high quality of spatially-and phenotypically-defined cells in situ.In this thesis,we have developed a robust technology with laser capture microdissection(LCM)using immunofluorescence to identify targeted phenotypes(immuno-LCM-RNAseq)in order to acquire finely resolved transcriptional programs of defined cells from snap-frozen or RNAlater-treated tissues at superb RNA qualities that are sufficient to resolve even isoforms.Here,we introduced the use of small-molecule RNase inhibitor,ribonucleoside vanadyl complex RVC,to prevent the long standing problem in RNA degradation during the lengthy process of immunofluorescence staining and microdissection of tissue sections under ambient conditions.In fact,the problem of RNA degradation has been a major obstacle in the widespread use of the capable LCM in demanding biological and biomedical questions.We have characterized the robustness of our technology with a variety of tissue types and under different storage conditions to show that high-quality RNA and excellent immunofluorescent images can be achieved simultaneously even with tissue types containing abundant endogenous RNases,enabling accurate acquisitions of spatial transcriptomes of as few as tens of defined cell types from tissues.Our results demonstrate that our technology outperformed existing methods by a large margin.Our technology is also compatible with tissues retained in RNAlater solutions,making it ideally suited for the investigation of clinical samples.As many tissue banks have adopted this method of preservation for the efficacy of RNA protection of RNAlater,our technology should further heighten the significance of these invaluable resources in the delineation of a number of debilitating human diseases,including cancer.To further demonstrate the power of our technology,we,for the first time,examined the capillary lymphatic tube in the small intestine villi,the lacteal,and revealed the genome wide expression differences of the tip and the tube cells at the transcript level.Although these cells are generally considered the same phenotype,functional differences have also been reported and the cells at the lacteal tip have clearly identifiable morphologies.Among the differentially expressed genes,it is quite interesting to find that the gene Tnfsf15 which is known to play a role in lymphatic endothelial cell growth and migration is specifically expressed in lacteal tip cells,whereas Arpc1a-201 isoform and Mtbp gene are not but robustly expressed in the tube cells,both of them encoding actin cutting proteins.These findings should provide us a solid basis for mechanistic examinations of the maintenance and growth of the lacteal whose abnormal functions are involved in several diseases.Combined with the extensive phenotype-specific antibodies accumulated over the past several decades,the technology developed in this thesis has significantly broadened the application of the powerful method of LCM to the rapidly advancing front of spatial genomics,and will provide a complementary approach to single cell(sc RNAseq)and/or coded beads based technologies(i.e.Slide-seq and HDST)where the true in vivo states of the defined cells can be obtained and compared.In fact,the actual transcriptome inferred from clusterings of sc RNA acquisitions must be verified or revised by their in vivo counterparts that could not be obtained until now.Even for the innovative coded beads based technology,the transcriptomes acquired with immuno-LCM-RNAseq may prove to be indispensable in the interpretation of the complicated datasets with these recent technologies.We anticipate that the robust immuno-LCM-RNAseq technology described in this thesis will become one of the major methodologies in biomedicine,especially for the delineation of developmental and disease processes,and undoubtedly make significant contributions to biology and the science of pathology.