Nat Genet: Resolving the effect of DNMT3A R882 mutation in human clonal hematopoiesis based on single-cell multiomics technology

Somatic mutations are prevalent in healthy tissues in humans, and their proportion also increases
with physiological age.
Somatic mutations in clonal growth products of healthy human tissue have been shown to overlap with cancer recurrence drivers (such as TP53, PIK3CA, and NOTCH1), suggesting that cancer may be caused
by precancerous clonal mosaicismic products.

Clone hematopoiesis (CH) usually refers to myeloid cell proliferation associated with an acquired gene mutation and is associated with
aging.
DNMT3A is the most frequently mutated gene in CH and encodes a DNA methyltransferase
.
Hotspot mutations in DNMT3A R882 are common in CH, are associated with an increased risk of myeloid malignancies, and are suspected to represent the early stages
of the development of certain tumors (such as acute myeloid leukemia).
CH provides a unique environment to study the molecular consequences of DNMT3A mutations in non-malignant human hematopoiesis, but the mutant cells in CH are morphologically and phenotypically similar to mixed wild-type cells, making it difficult for researchers to link
genotype to phenotype.

Recently, the team of Weill Cornell Medical College in the United States and the research team of Dana-Farber Cancer Institute published a report in Nature Genetics entitled "Single-cell multi-omics of human clonal hematopoiesis reveals that DNMT3A R882 mutations perturb early progenitor states.
" through selective hypomethylation"
.
The research team applied single-cell multi-omics sequencing technology to capture the mutational status of individual cells as well as downstream epigenetic and transcriptional information, and directly compared mutant cells from the same individual with their wild-type counterparts in primitive human samples
.
The study describes in detail the effect of DNMT3A R882 hotspot mutations on CH and the effect of somatic drive mutations on clonal chimeras, revealing that DNA methyltransferase mutations commonly found in blood stem cells may lead to disease through
changes in methylation and gene expression.

The article was published in Nature Genetics

Main research content

DNMT3A R882 genotyping in human CD34+ single-cell transcriptome sequencing

The research team first isolated live CD34+ cells from CH samples and genotyped transcriptome data using GoT30, capturing cells
with R882 codons.
To map the trajectory of CD34+ progenitor cell differentiation in CH, the research team integrated transcriptome data from different samples to determine the expected progenitor cell subtype
.
The results showed that DNMT3A mutations did not form new cellular features, consistent
with the fact that CH patients did not show significant peripheral blood counts or morphological abnormalities.
Instead, the research team observed that mutant and wild-type cells in CH were mixed together throughout the process, so more in-depth studies
were needed to link genotypes to cell phenotypes through single-cell multiomics.

Figure 1.
Mixed distribution
of DNMT3A R882 mutation and wild-type cells during clonal hematopoietic differentiation.
Source: Nature Genetics

DNMT3A-mutated human progenitor cells exhibit lineage bias

The research team mapped the above data to the hematopoietic differentiation pathway and found that cell differentiation with DNMT3A R882 hotspot mutation showed myeloid bias rather than lymphoid bias
.
In addition, mutated immature myeloid progenitors showed megakaryocyte-erythrocyte line expansion
.
The above results show that in the early stage of malignant transformation of hematopoietic system tumors, the concentration of transcription factor activity is accompanied by lineage and transcriptional disorders
.

Figure 2.
DNMT3A R882 mutant CH cells showed a pronounced tendency to
differentiate at key nodes.
Source: Nature Genetics

Multiomics to resolve the consequences of DNMT3A mutations

To correlate transcriptional regulatory network changes and DNA methylation characteristics caused by DNMT3A mutations, the team performed multimodal single-cell sequencing of CD34+ progenitor cells from the same individual to capture DNA methylation, transcriptome information, and targeted DNMT3A genotyping, allowing the methylation status of mutations from the same individual and wild-type cells to be compared, and single-cell transcriptome data determined the primary progenitor cell identity
.
It was found that the DNA methylation signature of the CpG island in the mutant group was reduced
.
In-depth analysis showed significant hypomethylation as well as global hypomethylation
in the enhancer region of DNMT3A R882 cells compared to wild-type cells.

Figure 3.
DNMT3A R882 promotes selective hypomethylation
of PRC2 targets during human hematopoiesis.
Source: Nature Genetics

Because DNMT3A encodes DNA methyltransferases, the team specifically examined the effects of
hotspot mutations on methylation patterns and gene expression.
The results showed that DNMT3A R882 hotspot mutations led to preferential hypomethylation
of polycomb inhibitory complex 2 (PRC2) targets as well as specific CpG flanking motifs.
Notably, hypomethylation motifs are rich in binding sequences of hematopoietic transcription factors such as MYC/MAX, HIF1A/ARNT, and USF1/2, which can serve as a potential association mechanism
between DNMT3A mutations and abnormal transcriptional phenotypes.
In addition, hypomethylation of the MYC/MAX allele also leads to increased
expression of its transcriptional targets.

The above results show that a single-cell multiomics approach provides a potential model for the analysis of transcriptional aberrations caused by DNMT3A mutations in CH, supporting the enhancement of the fitness
of DNMT3A mutant cells through selective hypomethylation of key hematopoietic transcription factor binding sequences.

Figure 4.
DNMT3A R882 shows sequence specificity
associated with MYC-binding motifs.
Source: Nature Genetics

Conclusion

Taken together, the study reported the molecular outcome of DNMT3A R882 mutations in human CH primary CD34+ cells and identified key epigenetic and transcriptional variants that remodel lineage biases of hematopoietic differentiation and promote clonal expansion in the earliest stages of tumor life
.
At the same time, the study also demonstrates the unique advantages of emerging single-cell multiomics approaches in solving complex scientific problems, paving the way
for studying how somatic mutations drive normal tissue chimeras in human somatic evolution.

Anna S.
Nam, Ph.
D.
, first author of the paper, said: "We have discovered a mechanism
that links DNMT3A R882 hotspot mutations to atypical transcriptional phenotypes.
These findings pave the way
for possible future interventions to target related cells to prevent cancer and other clonal growth-related diseases.
”

References:

Nam, A.
S.
, Dusaj, N.
, Izzo, F.
et al.
Single-cell multi-omics of human clonal hematopoiesis reveals that DNMT3A R882 mutations perturb early progenitor states through selective hypomethylation.
Nat Genet (2022).