FIN-Seq: Transcriptional profiling of specific cell types in frozen archived tissue from the human central nervous system

Thousands of frozen, archived tissues from postmortem human central nervous system (CNS) are currently available in brain banks. As single cell and single nucleus technologies are beginning to elucidate the cellular diversity present within the human CNS, it is becoming clear that transcriptional analysis of the human CNS requires cell type specificity. Single cell and single nucleus RNA profiling provide one avenue to decipher this heterogeneity. An alternative, complementary approach is to profile isolated, pre-defined cell types and use methods that can be applied to many archived human tissue samples. Here, we developed FIN-Seq (Frozen Immunolabeled Nuclei Sequencing), a method that accomplishes these goals. FIN-Seq uses immunohisto-chemical isolation of nuclei of specific cell types from frozen human tissue, followed by RNA-Sequencing. We applied this method to frozen postmortem samples of human cerebral cortex and retina and were able to identify transcripts, including low abundance transcripts, in specific cell types.


19
The human central nervous system (CNS) comprises an ex-20 tremely diverse set of cell types. While this heterogeneous cel-21 lular composition has been appreciated since the work of early  Transcriptional profiling of heterogeneous populations is 39 feasible with either single cell RNA sequencing (Macosko et al., 40 2015;Shekhar et al., 2016;Tasic et al., 2016;Zeisel et al., 2015) 41 or bulk RNA sequencing of purified user-defined cell types la-42 beled either genetically or with dyes and antibodies (Arlotta et 43 al., 2005;Heiman et al., 2008;Lobo et al., 2006;Molyneaux et 44 al., 2015;Siegert et al., 2012;Telley et al., 2016). Single cell 45 RNA sequencing has become essential for cataloguing molec-46 ularly distinct cell types in heterogeneous tissues such as the 47 CNS. However, sampling the whole tissue for rare cell types, 48 such as cone photoreceptors, is expensive as large numbers of 49 single cells need to be profiled. Alternatively, bulk RNA se-50 quencing of user-defined cell types allows for the acquisition of 51 transcriptomes of rarer cell types; thus, avoiding sequencing of a 52 large number of more abundant cell types. Acquisition of more 53 transcriptomes via single cell RNA sequencing is accelerating 54 the discovery of potential new markers that could be used to iso-55 late specific, rare cell populations from cellularly diverse tissues. 56 We aimed to develop a method that enables bulk RNA sequenc-57 ing of specific cell types and extends to archived frozen tissue. 58 Thousands of frozen human postmortem brain tissue samples, 59 including those with disease, are readily available through brain 60 banks, and they represent a crucial resource that is immediately 61 available and largely untapped. The abundance of archived CNS 62 tissue samples is crucial for profiling transcriptional changes in 63 rare diseases, and it is also likely that the number of biological 64 replicates needed in human studies is high because of the natural 65 genetic variation present among individuals.

160
From the DE analysis, we found an enrichment of known 161 CPN and Layer 4 (L4) markers (e.g. Cux2, Unc5d, and Rorb) in 162 the SATB2 + population among the unbiased top 50 DE genes. 163 Conversely, we found an enrichment of CFuPN markers (e.g. 164 Fezf2, Foxp2, and Crym) in the BCL11B + population (Figure 165 1g). BCL11B also labels interneurons in all layers of the mouse 166 neocortex (Arlotta et al., 2005;Nikouei et al., 2016). Accord-167 ingly, we found an enrichment of some interneuron markers in 168 the BCL11B + population (e.g. Gad1 and Gad2) (Figure 1g). To 169 confirm the molecular identities of the isolated neuronal pop-170 ulations, we also determined the relative expression levels of 171 known CPN and CFuPN marker genes that were differentially 172 expressed between CPN and CFuPN in previous studies (21 173 CPN markers and 22 CFuPN markers) (Arlotta et al., 2005;174 Molyneaux et al., 2007;Molyneaux et al., 2015). We found 175 that all CPN markers were enriched in the SATB2 + population 176 and all CFuPN markers were enriched in the BCL11B + popula-177 tion (Figure 1-figure supplement 2). To validate the differen-178 tially expressed genes, we chose four DE genes (Ddit4l, Unc5d, 179 Kcnn2, and Rprm) for further analysis. Using RNAscope dou-180 ble fluorescent in situ hybridization (FISH), we localized the 181 transcripts of these genes in specific neuronal populations. We 182 found that Ddit4l and Unc5d were expressed in layers 2 through 183 4 and were localized to SATB2 + neurons (Figure 1-figure sup-184 plement 3). Additionally, Kcnn2 and Rprm were expressed in 185 layers 5 and 6, respectively, and they were specifically confined 186 to BCL11B + neurons (Figure 1-figure supplement 3). In ad-187 dition, we successfully isolated and profiled the same neuronal 188 populations from mature, adult (1+ years old) mouse neocor-189 tex ( mouse brain. Nuclei were extracted from the frozen mouse neocortex by Dounce homogenization. The nuclei were fixed and immunolabeled with anti-BCL11B and anti-SATB2 antibodies. Two nuclear populations were isolated by FACS based on expression level of these two proteins. The nuclei were reverse crosslinked by protease digestion, and the RNA was extracted. Sequencing libraries were generated and subsequently sequenced to obtain cell type specific transcriptomes. (B) Representative immunohistochemistry images using BCL11B and SATB2 antibodies in the P30 mouse neocortex showed SATB2 expression in the upper layers and BCL11B expression in the deep layers (Left image). In layer 5, there were sparse cells that express both SATB2 and BCL11B (Bottom right image). (C) FACS plot of nuclei labeled with SATB2 and BCL11B antibodies showed a cluster of nuclei immunolabeled with BCL11B and a cluster of nuclei labeled with SATB2. (D) Isolated nuclei were counterstained by the Hoechst dye and either SATB2 or BCL11B in the SATB2 + population (top panels) or BCL11B + population (bottom panels). (E) Representative quantification of read counts mapped by transcript position (5' to 3') for every gene. (F) Representative quantification of percentage of read counts mapped to exonic, intronic, or intergenic regions of the genome. (G) Heatmap of unbiased top 50 differentially expressed genes between SATB2 + and BCL11B + populations. Known markers of callosal projection neurons (in red) were enriched in the SATB2 + population while known markers of corticofugal projection neurons (in green) were enriched in the BCL11B + population. Scale bars; 100 µm (b, right panels, d), 500 µm (b, left panel).
To determine the degree to which nuclear transcript abun-  To determine the molecular identity of the SATB2 + and 255 BCL11B + populations, we first compared the gene expression 256 levels of known markers of oligodendrocytes, astrocytes, and 257 neurons. We found that neuronal markers were enriched in both 258 SATB2 + and BCL11B + populations. We also found an enrich-259 ment in the BCL11B + population of PDGFRA, normally con-260 sidered an oligodendrocyte marker, but also previously shown (also known as VGLUT1) and did not express GAD1 or GAD2, 265 while the BCL11B + population expressed GAD1 and GAD2 at 266 high levels, indicating that, while SATB2 + population contained 267 mainly excitatory neurons, BCL11B + population contained also 268 inhibitory neurons (Figure 2-figure supplement 4). 269 We next sought to understand the identity of the SATB2 + and 270 BCL11B + populations at the neuronal subtype-level. Previously, 271 single nucleus RNA-seq has identified eight excitatory neuronal 272 subtypes (Ex1-Ex8) and eight inhibitory neuronal subtypes (In1-273 In8) in the adult human neocortex (Lake et al., 2016). SATB2 274 is expressed in all excitatory neurons, but it is most highly ex-275 pressed in one of the neuronal subtypes referred to, in this prior 276 study, as Ex4. BCL11B is highly expressed in In1, In4, In5, and 277 In6. SATB2 and BCL11B are both expressed in Ex6 and Ex8, 278 but we would not expect to see these subtypes in our popula-279 tions as we did not collect the SATB2 HI BCL11B HI population. 280 For the SATB2 + population, we cross-referenced our DE gene 281 set (adjusted p-value < 0.05) to the molecular signature genes 282 that define the eight excitatory cortical neuronal subtypes (Ex1-283 Ex8). From this analysis, we observed a high level of expres-284 sion of Ex4 markers in the SATB2 + population compared to the 285 All population (Figure 2c). To confirm these results, we also 286 ran the dataset through a gene set enrichment analysis (GSEA) 287 against all marker genes that define Ex1-Ex8 (Subramanian et 288 al., 2005). We found that Ex4 gene set was significantly en-289 riched in the SATB2 + population while Ex6 and Ex8 gene sets 290 were enriched in the All population (default significance at FDR 291 < 0.25; Ex4: FDR = 0.139; Ex6: FDR = 0.043; Ex8: FDR = 292 0.005). Depletion of Ex6 and Ex8 from the SATB2 + popula-293 tion is likely due to the exclusion of SATB2 HI BCL11B HI nuclei. 294 We confirmed the expression of COL6A1 and ANXA1, two Ex4 295 markers, in SATB2 + neurons by single molecule FISH (Figure 296 2d). In the BCL11B + population, we found that the markers 297 for In1, In4, In5, and In6 were enriched compared to the All 298 population (Figure 2e). Furthermore, previous single cell se-299 quencing of the fresh adult human brain identified seven neu-300 ronal communities (NC), of which SATB2 is highly expressed in 301 neuronal community 4 (NC4) (Darmanis et al., 2015). Accord-302 ingly, we found that the markers for NC4 are highly expressed 303 in the isolated SATB2 + population (Figure 2-figure supple-304 ment 5). By GSEA analysis, we also found that NC4 gene 305 set was significantly enriched in the SATB2 + population (FDR 306 = 0.037). Taken together, our results show the FIN-Seq proto-307 col can isolate molecularly-defined neuronal subtypes for down-308 stream transcriptional profiling from frozen postmortem human 309 cortical samples. To determine whether we could use FIN-Seq to isolate and pro-313 file specific cell types from another region of the human CNS, 314 we chose to isolate cone photoreceptors from the retina. We 315 obtained four frozen postmortem eyes (age range: 40-60, see 316 Materials and methods for description of samples) from patients 317 without known retinal disorders. Nuclei were extracted from 318 the mid-peripheral retina, fixed, and immunostained by a hu-319 man Cone Arrestin (CAR, also known as ARR3) antibody (Fig-320  ure 3a). In human retinal cross-sections, we found CAR expres-321 sion in the nuclei and cell bodies of cone photoreceptors, located 322 in the outer nuclear layer where all photoreceptors reside (Fig-323  Nuclei were isolated and subsequently fixed in 4% PFA. They were immunolabeled with anti-BCL11B and anti-SATB2 antibodies, and FACS isolated into populations. RNA from the nuclei were sequenced to obtain a cell type specific transcriptome. (B) Representative immunohistochemistry of the adult human cerebral cortex using anti-BCL11B and anti-SATB2 antibodies. Some nuclei expressed both SATB2 and BCL11B (arrows), some nuclei expressed BCL11B but not SATB2 (arrowheads), and many nuclei expressed SATB2 but not BCL11B. (C) A heatmap representing relative expression levels of excitatory neuron markers previously identified by single nuclei RNA sequencing that are differentially expressed (adjusted p-value<0.05) between SATB2 + and All populations. Markers of neuronal subtype Ex4 (outlined in red), which expresses SATB2, were enriched in the SATB2 + population. (D) Validation of Ex4 markers, COL6A1 (left panel) and ANXA1 (right panel) using RNAscope single molecule FISH. Both COL6A1 and ANXA1 were expressed in SATB2 + neurons (arrows). (E) A heatmap representing relative expression levels of inhibitory neuron markers previously identified by single nuclei RNA sequencing that are differentially expressed (adjusted p-value<0.05) between BCL11B + and All populations. Markers of neuronal subtypes, In1, In4, In5, and In6, all of which express BCL11B, were enriched in the BCL11B + population. Scale bars: 100 µm (b), 50 µm (d).
ure 3b). CAR + and CARnuclei were isolated by FACS, and Seq v.4 and sequenced to a mean depth of 43 million (range: 37 336 53 million reads/replicate) 75bp paired-end reads. The sequenc-337 ing reads were analyzed, and the quality control parameters in-338 dicated successful sample separation and differential expression 339 analysis (Figure 3-figure supplement 2). We found 5,260 DE 340 genes (adjusted p-value < 0.05) between CAR + and CARnu-341 clear populations.

342
To determine the cellular identity of the CAR + population, 343 we examined the top 50 differential expressed genes between 344 CAR + and CARpopulations. Of the 16 genes enriched in the 345  (Bossers et al., 2009;Dangond et al., 2004;Dumitriu et al., 2012;Hauser et al., 2005;Hawrylycz et al., approaches cannot account for cellular heterogeneity of the hu-373 man brain, an organ with tremendous cellular diversity. This is  Mitchell and Borasio, 2007;Sulzer and Surmeier, 2013).

378
To understand the transcriptional changes that accompany cel-   (Venkataraman et al., 417 2018). For molecular markers without an antibody, FIN-Seq 418 could be further developed to isolate specific cell populations 419 using nuclear RNA by FISH techniques such as RNAscope or 420 SABER (Kishi et al., 2018;Klemm et al., 2014). Labeling 421 specific nuclear transcripts of human neuronal nuclei for down-422 stream FACS and transcriptome sequencing will enable FIN-Seq 423 to capture any cell type of interest.

487
Sorted nuclei were spun at 3000xg for 7 minutes, and the supernatant was dis-   (Okonechnikov et al., 2016). Read counts were generated by HT-seq ver-511 sion 0.6.1 (Anders et al., 2015). Sample parameters used were as follows: -i 512 gene name -s no 513 The resulting matrix of read counts were analyzed for differential expres- antibodies. FFPE adult human cerebral cortex tissue from a 54-year-old female 552 was obtained from Abcam (ab4296). The brain tissue was deparaffinized by 2x 553 xylene incubation (3 minutes each) followed by 1x 100% ethanol (3 minutes), 554 1x 95% ethanol (3 minutes), 1x 70% ethanol (3 minutes) washes. Antigen re-555 trieval was performed in a citrate buffer (10mM Citric Acid, pH 6.0) in a rice 556 cooker with boiling water for 20 minutes. Subsequently, immunohistochemistry 557 was performed as described above with an additional step of incubation in True-558 Black (Biotium, Fremont, CA) after incubation in blocking buffer to quench the 559 lipofuscin autofluorescence. For human eye immunohistochemistry, formalin-560 fixed human postmortem eyes were obtained from Restore Life USA. Patient 561 DRLU101818C is a 54-year-old male with no clinical eye diagnosis and the 562 postmortem interval was 4 hours. Patient DRLU110118A is a 59-year-old fe-563 male with no clinical eye diagnosis and the postmortem interval was 4 hours. 564 The retina was cryosectioned at 16 µm thickness. Immunohistochemistry was 565 performed as previously described (Arlotta et al., 2005) with anti-CAR antibody 566 (1:10,000).