Efficient and sensitive profiling of human BCR repertoire
Next generation sequencing (NGS) approaches for profiling human B-cell receptors (BCRs) yield powerful insights into the immune response of healthy individuals and those affected by diseases. To fully harness the potential of such profiling, sensitive, reliable techniques to amplify BCR repertoires are needed.
SMART-Seq Human BCR (with UMIs) (referred to as “hBCRv2”) can reproducibly detect all BCR clonotypes, even those with low abundance, with best-in-class sensitivity. Using a high-throughput mRNA-based method for human BCR profiling, you can amplify variable regions of all BCR isotypes, both heavy and light chains.
B cells are vital members of the adaptive humoral immune response. They express receptors (BCRs) on their surface. BCRs recognize the unique molecular patterns of invading pathogens and block the spread of infection by secreting neutralizing antibodies. A developing B cell undergoes recombination; afterward, the BCR has two identical heavy chains (generated by the recombination of V, D, and J gene segments) and two identical light chains (generated by the recombination of V and J gene segments).
The diverse collection of clonotypes—the BCR repertoire—is key to shaping immunity. Profiling this repertoire with NGS has provided valuable insights, such as informing the design of therapeutic interventions against SARS-CoV-2. Current technologies are limited in their ability to generate data accurately and reproducibly for all human BCR heavy and light chains. There is an urgent need for both academic and clinical researchers to get a precise readout of BCR sequences for a good understanding of B cell responses in normal and disease states.
SMART-Seq Human BCR (with UMIs) kit chemistry
We developed a refined kit for profiling the BCR repertoire to satisfy this need. It leverages SMART technology (Switching Mechanism at 5′ End of RNA Template) and pairs NGS with a 5′ RACE approach. This optimized chemistry for BCR profiling delivers sensitivity, accuracy, and reliability. The 5′-RACE method used in hBCRv2 does not require any prior knowledge of the BCR sequences, reduces variability, and allows for priming from the constant region of all heavy and light chains (Figure 1).
Figure 1. SMART-Seq Human BCR (with UMIs) workflow. dT-primed, first-strand cDNA synthesis is followed by two rounds of successive PCR for amplification of cDNA sequences. After post-PCR purification, size selection, and quality analysis, the library is ready for sequencing.
When used with CogentIP, PCR duplicates and sequencing errors can be detected and removed from the data, leading to more accurate and reliable clonotype calling and quantification. The addition of the UDIs lets researchers pool multiple samples—currently up to 384—while providing greater confidence in sample integrity when sequencing on a patterned Illumina flow cell. We have improved our chemistry from our initial human BCR kit (hBCRv1) to make the BCR libraries compatible with any Illumina sequencer, allowing researchers to save on sequencing costs and increase sample multiplexing on high-throughput sequencers such as the NovaSeq™ system. Full-length V(D)J transcript information can be obtained when libraries are sequenced on the MiSeq® or NextSeq® system, or the CDR3 region using shorter sequencing read lengths on any sequencer. hBCRv2 is designed to generate a consistent library yield from 10 ng–1 ug of PBMC RNA, 1 ng–100 ng of B cell RNA, or up to 100 ng of whole blood RNA, ensuring a sufficient yield for any sequencing configurations. We have also deployed pooled primers to greatly simplify the workflow and allow researchers to study heavy chains found in all isotypes (IgG, IgM, IgD, IgA, and IgE) and immunoglobulin light chain (IgK and IgL) in a much more seamless way.
Results
More sensitive and comprehensive BCR profiling with hBCRv2
To evaluate the sensitivity of hBCRv2, libraries were prepared from 10, 100, and 1,000 ng of RNA extracted from PBMCs, generating approximately 2 million, 6 million, and 12 million sequencing reads, respectively. Clonotype counts from both heavy chain (IgG/M/D/A/E) and light chain (IgK and IgL) consistently increase as the amount of RNA input increases (Figure 2, Panel A). Chord diagrams for both heavy-chain (HC) and light-chain (LC) libraries prepared from 10 ng input showed confident detection of clonotypes with varying abundance, as indicated by chord of various thickness (Figure 2, Panel B).
Figure 2. hBCRv2 detected all BCR clonotypes with confidence.Panel A. Detection of all BCR clonotypes for both heavy chain (IgA/D/E/G/M) and light chain (IgK/L) by hBCRv2 in 10, 100, and 1,000 ng input of PBMC RNA. Comparison across various RNA inputs shows consistent increase in clonotype count with increase in RNA input. Panel B. Chord diagram for heavy-chain (HC) and light-chain (LC) isotypes of 10 ng input of PBMC RNA.
To demonstrate how SMART technology combined with 5′ RACE is the best-in-class approach for BCR profiling, we prepared BCR libraries using 10 ng of human PBMC total RNA according to the manufacturer’s instructions. With sequencing reads downsampled to ~1 million for Company N libraries and 500,000 each for heavy- and light-chain libraries for hBCRv2, we found a 127% increase in clonotype counts against Company N, manifesting superior data quality (Figure 3).
Figure 3. Best-in-class sensitivity of hBCRv2. Libraries were generated using 10 ng of human PBMC total RNA according to the manufacturer’s instructions. The resulting cDNA libraries were sequenced on Illumina platform. Sequencing outputs were downsampled to ~1 million for Company N libraries and 500,000 each for heavy and light chain libraries for hBCRv2. In comparison to Company N, hBCRv2 generated approximately 9.5K and 22.9K clonotypes for heavy-chain (HC) and light-chain (LC) isotypes respectively, representing a 127% increase against Company N.
A streamlined workflow to profile all heavy chains
One of the most important improvements we made in hBCRv2 as compared to hBCRv1 was a much more simplified workflow enabling detection of all BCR heavy chains. While the previous kit only allowed researchers to sequence IgG and IgM for the heavy chains, hBCRv2 can profile all BCR heavy chains (i.e., IgG/M/D/A/E) with great confidence. Compared with hBCRv1, hBCRv2 also displays much higher sensitivity, as indicated by more clonotypes detected for both heavy and light chains at the same sequencing depth (Figure 4). Besides revamped chemistry to support ultra-sensitive readout of more BCR clonotypes, we also redesigned the workflow with pooled primers, allowing researchers to generate all BCR heavy-chain libraries in one reaction.
Figure 4. More clonotypes captured by a simpler workflow. Libraries were generated from 1 ng of B-cell RNA using the hBCRv1 (light blue) and hBCRv2 (dark blue) workflow. The resulting cDNA libraries were sequenced on an Illumina MiSeq sequencer. Sequencing data was downsampled to 250,000 for each heavy and light chain of hBCRv1 libraries and 500,000 for each heavy and light chain of hBCRv2 libraries. Data was analyzed using CogentIP. Here, the bar graphs illustrate that higher numbers of and more diverse clonotypes were detected with the improved hBCRv2 chemistry.
hBCRv2 provides accurate and unbiased amplification of low-abundance clonotypes
Another key benefit of hBCRv2 is its superior accuracy powered by UMIs. To evaluate how UMIs guaranteed accurate profiling of BCR clonotypes, we spiked 100, 10, 1, and 0.1 pg of RNA extracted from the TIB-190 cell line into 10 ng of PBMC RNA. BCR libraries were prepared with hBCRv2 and sequenced on an Illumina MiSeq sequencer. During data analysis, libraries were normalized to 200,000 reads and clonotype counts were measured after UMI-based consensus collapse in CogentIP. The ratio of detected TIB-190 reads versus total reads showed good correlation with the spike-in percentage of TIB-190 RNA (Figure 5). This result indicates that UMI-powered analysis with hBCRv2 enables researchers to identify low-abundance BCR clonotypes with great confidence and accuracy.
Figure 5. Successful identification of low-abundance clonotypes. 100, 10, 1, and 0.1 pg of RNA extracted from the TIB-190 cell line were spiked into 10 ng of PBMC RNA. Panel A. Clone counts at different spike-in levels are listed in the table. Libraries were normalized to 200,000 reads, and all counts were measured after UMI-based consensus collapse. Panel B. Calculated correlation between spike-in RNA proportions and detected clonotype frequencies. Both axes are logarithmically transformed.
Best-in-class reproducibility is needed for biomarker discoveries in clinical research
Immune repertoire sequencing, such as BCR profiling, has been widely used for biomarker discoveries in clinical research, making it extremely important that the underlying workflow can generate reproducible and high-quality data each time. To evaluate the reproducibility of hBCRv2, we prepared BCR libraries from 10 ng of PBMC RNA and 1 ng of B-cell RNA using the hBCRv2 protocol. Technical replicates prepared with 10 ng of PBMC RNA were sequenced on a MiSeq sequencer with 300-cycle and 600-cycle kits. The Venn diagram illustrates 87% clonotype overlap (Figure 6, Panel A). Libraries prepared using B-cell RNA also displayed great reproducibility, as indicated by ≥85% clonotype overlap between the libraries sequenced on different platforms (Figure 6, Panel B).
Figure 6. Evaluation of clonotype reproducibility. BCR profiling libraries from 10 ng of PBMC RNA and 1 ng of B-cell RNA were prepared using the hBCRv2 workflow. Panel A. Technical replicates prepared with 10 ng PBMC RNA were sequenced on the Illumina MiSeq platform. Data generated were downsampled to 1 million reads and analyzed using CogentIP. The Venn diagram illustrates 87% clonotype overlap between technical replicates libraries. Panel B. Libraries prepared using 1 ng B-cell RNA were sequenced on the same MiSeq sequencer with both a 600-cycle V3 cartridge and a 300-cycle V2 cartridge, as well as on the Illumina MiniSeq™ platform with a 300-cycle cartridge. Data generated were downsampled to 1 million reads and analyzed using CogentIP. Venn diagrams show ≥85% clonotype overlap between the libraries sequenced on different platforms.
Conclusion
SMART-Seq Human BCR (with UMIs) is an innovative and powerful tool for profiling human B-cell receptors. By leveraging SMART technology and combining a 5′-RACE approach with gene-specific amplification, this workflow captures complete V(D)J variable regions of BCRs and is optimized for sensitive and reproducible clonotype detection. Incorporation of UMIs during reverse transcription also enhances the accuracy of this workflow by removing sequencing reads derived from PCR amplification bias. Furthermore, we have also modified the workflow with pooled primers so researchers can enjoy a much easier protocol with all BCR chains profiled in one reaction. Incorporating UDIs into the libraries allows for both pooling of up to 384 samples and sequencing on patterned flow cells without worrying about index hopping. Lastly, we provide researchers with a complete end-to-end solution, as our Cogent NGS Immune Profiler Software and Cogent NGS Immune Viewer feature an easy-to-use method for analysis and visualization of the immune repertoire sequence information at your desktop.
Materials and Methods
BCR libraries were prepared from human PBMC total RNA (Takara Bio, Cat. # 636592) and human B-Cell (CD19+) total RNA (Miltenyi Biotech, Cat. # 130-0930169). For the sensitivity test, total RNA from the TIB-190 B-cell carcinoma cell line (ATCC) was spiked into 10 ng of PBMC RNA at the listed concentrations. Total RNA from TIB-190 cells was extracted using the NucleoSpin RNA plus kit (Takara Bio, Cat. # 740984.50).
All BCR libraries were generated using SMART-Seq Human BCR (with UMIs), as per the user manual. Following purification and size selection, libraries were quantified using the Qubit and Agilent Bioanalyzer 2100. Sequencing was performed on a MiSeq sequencer with 600-cycle V3 cartridges (Illumina, Cat. # MS-102-3003), a MiSeq sequencer with 300-cycle V3 cartridges (Illumina, Cat. # MS-102-3001), and a MiniSeq sequencer with 300-cycle cartridges (Illumina, Cat. # FC-420-1002). Data analysis was completed using the Cogent NGS Immune Profiler Software. The report of heavy-chain and light-chain clonotypes from 10 ng of PBMC RNA generated by CogentIP was uploaded to VDJviz browser for chord diagram visualization.
