Variant load of Mitochondrial DNA in single human Mesenchymal stem cells

doi:10.21203/rs.3.rs-3735071/v1

Download PDF

Article

Variant load of Mitochondrial DNA in single human Mesenchymal stem cells

https://doi.org/10.21203/rs.3.rs-3735071/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Mesenchymal stem cells (MSCs) are the parent cells to many cells of the musculoskeletal system including osteoblasts. Previous work has shown that in mouse models mitochondrial DNA (mtDNA) pathological variants lead to dysfunction of the respiratory chain of these osteoblasts and the premature development of age related osteoporosis. An increased rate of respiratory chain deficiency has also been observed in human osteoblasts.

An experimental pipeline to isolate single MSCs using flow cytometry was developed before performing next generation sequencing to analyse the pathogenic variant load of 13 patients aged 22-88. In all patients, somatic mtDNA variants were detected in individual cells. As per previous studies, increased variants within the mtDNA control region (D-Loop) were detected. However overall, there was no significant difference in the distribution of variants across the rest of the genome in all patients. Although a higher proportion of non-synonymous variants were seen this was not statistically significant.

It is possible to isolate and sequence individual MSCs and detect somatic mtDNA variants. This gives a snapshot in time of the variant load for each patient. Due to low sample numbers when compared to studies investigating the role of pathogenic mtDNA variants in other cell types, it was not possible to observe any age related increase in pathological variants. However, this this study adds further evidence that clonally expanded, somatic mtDNA variants are common and could contribute to age related diseases including osteoporosis.

Biological sciences/Biotechnology/Biomaterials/Biomaterials cells

Biological sciences/Biotechnology/Biomaterials/Dna and rna

Biological sciences/Cell biology/Mechanisms of disease

Biological sciences/Cell biology/Organelles

Biological sciences/Biological techniques/Sequencing/Dna sequencing

Biological sciences/Biological techniques/Sequencing/Next generation sequencing

Mesenchymal stem cells (MSCs) are parent cells to many cells of the musculoskeletal system and are important in the formation and repair of bone tissue. Differentiated cells include osteoblasts, chondrocytes, adipocytes and myocytes. With age, human tissue has been shown to acquire mitochondrial DNA (mtDNA) mutations which lead to defective respiratory chain function [1]. mtDNA has approximately 10 times the mutation rate of nuclear DNA, likely due to replication errors mediated by mtDNA polymerase [2]. These somatic pathogenic variants have been linked to the ageing phenotype with many systemic effects upon the musculoskeletal system [3].

Primary osteoporosis is part of the ageing profile and is characterised by reduced bone mass and microarchitecture deterioration. This results in a loss of strength leading to increased risk of fragility fractures with advancing age [4–6]. Bone mass and mineral density peak in the third decade before beginning to decline with advancing age [7]. The underlying mechanisms of osteoporosis are complex and multifactorial, but data from mice have suggested a potential role for mtDNA pathogenic variants [3, 8].

Investigating whether mtDNA variants have a role in age related disease is challenging. Mitochondrial DNA exists as multiple copies within each cell [9]. Pathogenic variants can be present in all copies (homoplasmy) or only in a proportion of copies (heteroplasmy) [10]. In the presence of heteroplasmy, there is a threshold level of pathogenic variant within the cell before a biochemical defect is seen. This threshold level can vary but is typically between 60 and 90% [11] for mtDNA single nucleotide polymorphic (SNP) variants. A SNP can reach high levels within cells due to clonal expansion [12]. Another factor complicating the investigation of the role of somatic mtDNA pathogenic variants in ageing, is that the individual variant will differ between cells since these somatic variants are random throughout the genome [13]. This means that to be confident of understanding the role of pathogenic variants of mtDNA in age related disease, identification of the variants must be investigated in individual cells.

Previous studies [8] revealed the presence of mitochondrial deficiency, due to mtDNA pathogenic variants in osteoblasts from mice, lead to accelerated bone loss. An increase in human osteoblasts with a respiratory chain deficiency seen with age has also been reported [14]. We wished to understand if mtDNA pathogenic variants could contribute to the ageing phenotype seen in human bone. Specifically, we wished to determine if there was any evidence of somatic mtDNA pathogenic variants in mesenchymal stem cells, the precursors to human osteoblasts. This could explain the increase in respiratory deficient osteoblasts seen in previous studies and thus play a role in the pathogenesis of osteoporosis.

Sample Cohort and MSC isolation

Bone marrow samples were taken at the time of elective hip arthroplasty and other routine orthopaedic surgeries from 13 patients (22–88 years old). Samples were collected from patients undergoing elective orthopaedic surgery from the Newcastle upon Tyne Hospitals. Ethical approval for use of adult human samples was gained as an adjunct to the Newcastle Bone and Joint Biobank – (REC reference 14/NE/1212, IRAS project ID166522). Newcastle University reference 8741/2016. All patients gave informed consent to the use of tissue as part of the ethical approval requirements and experiments were performed in accordance with relevant guidelines and regulations. Patient information is summarised in First mononuclear cells were concentrated and separated from the rest of the marrow using a density gradient (a 1.077 g/ml Lymphoprep™ density gradient medium (StemCell Technologies) [15]. This removed contaminants from the sample such as red cells, fat and multinucleated cells which were not of interest, MSCs were contained within an opaque band mononuclear layer at the Lymphoprep™ - marrow interphase. By concentrating MSCs within the mononuclear cells led to less demand on the flow cytometer in terms of number of cells to sort improving efficiency and the time taken.

Flow cytometry can be used to isolate individual cells [16]: here we used it to sort single MSCs from the mononuclear cell group. MSCs were identified using positive lineage fluorescent antibodies for CD105, CD90 and CD73 and negative lineage antibodies for CD14, CD20, CD34 and CD45. These markers were based on a Miltenyi Biotect kit and previous published work [17, 18]. Cells meeting the positive lineage markers without the negative markers were sorted into a single wells of a 96 well PCR plate using electrostatic deflection. As each MSC is identified on the basis of its fluorescent profile the stream within which the cell is contained is diverted to fall into a single well of a 96 well plate. The plate carriage then moves to align a new well for the next MSC to be diverted into within the stream. Due to the low frequency of cells this would take between 5 and 20 minutes to fill a plate. Within each well of the plate there was 15µl of single cell lysis buffer (0.5 M Tris-HCl, 0.5% Tween 20, 1% Proteinase K, pH 8.5). Following cell sorting, the plate was briefly spun at 150G for 2 minutes to ensure the cell was at the bottom of the well and within the lysis buffer.

Cell lysis was performed with an incubation period at 55°C for three hours, followed by 10 minutes at 95°C. The consensus sample (1 million cells isolated prior to flow cytometry preparation) was also lysed in preparation for PCR amplification.

Mitochondrial DNA Isolation and Deep Sequencing.

As in previous work [19], mtDNA was enriched using five overlapping long-range PCR amplicons to cover the entire mtDNA genome (Set 1 m.323–343 and m.3574 − 3556), (Set 2 m.3017–3036 and m.6944 − 6924), (Set 3 m.6358–6377 and m.10147 − 10128), (Set 4 m.9607–9627 and m.13859 − 13839) and (Set 5 m.13365–13383 and m.771 − 752). Amplified products were assessed by gel electrophoresis, against DNA + ve and DNA − ve controls, and quantified using a Qubit 2.0 fluorimeter (Life Technologies, Paisley, UK). Each amplicon was individually purified using Agencourt AMPure XP beads (Beckman-Coulter, USA), pooled in equimolar concentrations and re-quantified. Pooled amplicons were tagmented, amplified, cleaned, normalized, and pooled into 48 sample multiplexes using the Illumina Nextera XT DNA sample preparation kit (Illumina, CA, USA). Multiplex pools were sequenced using MiSeq Reagent Kit v3.0 (Illumina, CA, USA) in paired-end, 250 bp reads.

Bioinformatic Analysis

Postrun FASTQ files were analysed using an in-house bioinformatic pipeline. Briefly, sequence reads were quality check using FastQC (v. 011.8), mapped against genome version GRCh38, using BWA (invoking –mem, v. 0.7.17) [20] sorted and indexed using Samtools (v.1.12) [21]. All duplicated reads from the resulting bam files were marked with using Picard (v.2.2.4). mtDNA variants (mtSNVs) were called using bcftools (v1.10.2) and Mutserve (v.2.0.0) accordingly [22, 23]. mtSNVs with base quality (--baseQ) over 30 and minimum heteroplasmy level (--level) 0.02 were called. Low-quality variants, present in low-complexity regions [24] were not included in comparative analysis (e.g. 66–71 bp, 300–316 bp, 513–525 bp, 3106–3107 bp, 12418–21425 bp, 16181–16194 bp). Remaining variants were annotated via ENSEMBL VEP v107 [25]. Homoplasmic variation was defined as > 98% Heteroplasmic variation was defined as > 2% and < 98%. Heteroplasmic counts are based upon number of heteroplasmies per cell. Heteroplasmic Fraction (HF) is the proportion of mtDNA variant relative to reference.

Single Cell Quality Control

mtDNA Variant data was compared between single cells and a sample consensus sequence obtained from sequencing a sample of the cells from the mononuclear separated layer prior to flow cytometry sorting (10⁶ cells). Single cell data was assessed for homology to consensus using homoplasmic mtDNA variants (homology > 98%), coverage (> 95% coverage at > 1000 reads) and for cross-contamination between samples. After QC 99 cells across 13 samples remained.

Statistical analysis

Data was analysed using R using data appropriate tests (detailed in the text). Statistical significance was set at p < 0.05.

MSC identification and isolation

The combined use of a density gradient, flow cytometry single cell sorting and NGS was used to sequence individual MSCs. We have been able to isolate and sequence a limited sample of MSCs (99 single cells (after filtering and quality control steps) and 13 consensus samples) from 13 patients across an age range 22–88 years (8 females, 5 males).

Variant Data

Analysis of 99 single cells across 13 samples identified 9032 heteroplasmic variants (> 0%/<100%). On inspection, the vast majority (69%) were less than 2% (Fig. 1). For further analysis, and similar to previous work [26] we set a conservative threshold of 2%, selecting 3864 variants for further analysis (Fig. 1).

Although there was some variability between cells within the same tissue sample due to cell-specific clonal expansion, the distribution of heteroplasmic variant counts were highly similar between tissue samples (Dunnett’s p > 0.05, Fig. 2). Additionally, although specific cells harboured comparatively high-fraction variants (81 variants with heteroplasmic fraction between 50 and 98% for 12 out of 13 tissue samples), the overall distribution of heteroplasmic fractions was highly similar between tissue samples (Dunnett’s p > 0.05) and was typically below 10% (Fig. 2). Further analysis of heteroplasmic variants showed 31.94% of all mtDNA variants were T > C substitutions, 28.01% were A > G, 18.62% were C > T and 13.88% were G > A. The remaining possible polymorphisms (A > C, A > T, C > A, C > G, G > T, G > C, T > A and T > G) represented less than 2.25% of the variants detected.

Next, we investigated whether heteroplasmic variation was regional. Previous studies have shown increased mtDNA pathogenic variant rates in the mtDNA control region (D-Loop, m.16024-576, [27]). In line with this we observed a significant overrepresentation of heteroplasmic variants in the D-Loop compared to other regions of the genome when grouped (Dunnett’s p > 0.05, Fig. 3A). However, we observed no significant increase heteroplasmy fractions when stratified by locus type (Fig. 3B), indicating that although heteroplasmies are regionally correlated, their proportions are relatively similar between locus types and thus randomly distributed.

Finally comparing protein coding regions (Fig. 4) we found a higher proportion of non-synonymous variants per cell although there was no statistical difference in the heteroplasmic fraction distributions between non-synonymous and synonymous variants across the cohort.

Data Availability

The single cell dataset and the informatic pipeline used in the current study are publicly accessible in the Zenodo.org repository. Raw sequencing data (FASTQ files and Sample Manifest) is available at https://zenodo.org/records/10399925 (DOI: 10.5281/zenodo.10399925) and informatic pipeline used to generate mtDNA variant calls is available at https://zenodo.org/records/1157051 (DOI:10.5281/zenodo.1157050).

Here we show that it is possible to isolate single MSCs and sequence the mtDNA within these cells. There were several challenges to the completion and interpretation of this work. MSCs are scarce within the bone marrow they make up only 0.01% of the cell population in marrow [28]. Collection and single cell sorting of cells causes trauma to the cells and this could lead to damage and reduced coverage during sequencing of the mitochondrial genome. It is also a small amount of mtDNA to be amplified and analysed. Poor coverage and concern regarding artefacts led to the exclusion of 31 cells. Sequencing errors were typically found in areas of low coverage.

Tracking clonal expansion in MSCs is not possible in the same way that it is in colonocyte stem cells which are based at the bottom of a colonic crypt, and all cells derived from the stem cells migrate up crypt wall. This makes the colon ideal for linking somatic mtDNA variants with phenotypic changes in daughter cells [29]. In contrast MSCs respond to multiple different stimuli and differentiate into multiple cell lineages, such as bone, cartilage, adipose, muscle, tendon and stroma, or to self-renew [30]. Tracking clonal expansion from one cell to daughter cells from in vivo samples is therefore not possible. Thus, we do not know whether variants detected in MSCs play a role in respiratory deficiency seen in human osteoblasts but it would be logical to conclude this could be the case.

The results of this study give a snapshot into the potential pathogenic variant load of one sample of cells in comparison to a consensus sample of each patient at one specific time point. The use of the consensus sequence allowed identification and exclusion of inherited variants and the ability to identify clonally expanded somatic pathogenic variants.

The data has shown heteroplasmic mtDNA pathogenic variants do exist in MSCs; no somatic variant within the samples reached homoplasmic (100%) levels. The sample size of patients and cells sequenced was limited and does not allow determination of whether mtDNA pathogenic variants accumulate with age in MSCs, although the presence of respiratory chain deficiency has been seen with increasing age in human osteoblasts [14]. However, it is clear that somatic mtDNA variants do occur within single cells whatever the age of the patient (the youngest been 22). Previous work which compared iPS-MSCs (induced pluripotent stem cell -MSCs) to dental tissue-derived MSCs identified some mtDNA pathogenic variants in MSCs, although the full mitochondrial genome was not investigated in the same detail as in this study.

Although an increased frequency of mtDNA pathogenic variants and respiratory deficiency was seen with increasing age in colonocytes [31], this required a large number of samples from 207 patients aged 17–78, much greater than the samples analysed here. This work also demonstrated the presence of pathogenic mtDNA mutations from early adulthood or less than 20 years of age both at low levels and clonally expanded. They demonstrated that there is a low background level of mutation that is constant but that with age the number of clonally expanded mutations increases with age.

Similarly age related mtDNA variants in the colon also did not show statistical significant variance across the genome [32]. There was also an increased number of non-synonymous changes seen with age. Although greater non-synonymous changes were observed our data set is too small to correlate this with age or show statistical significance. A larger study would be needed to clarify if there is an age-related effect in terms of pathogenic variant load and in-vitro study to see how these pathogenic variants are passed onto daughter cells of different lineage.

It is possible to isolate and sequence single MSCs mitochondrial genome for analysis. The use of flow cytometry and methods here highlight a technique that could be adapted to any cell within suspension. Pathogenic variants were seen across the genome and with clear evidence of somatic mtDNA pathogenic variants. An overall trend in relation to age was not feasible to judge but adds further evidence that clonally expanded, somatic mtDNA variants are common and could contribute to age related disease. A larger study would be needed to clarify if there is an age-related affect as has been seen in other tissues and whether this has implications in the age-related development of osteoporosis.

Author Contribution

DH wrote the manuscript and carried out the majority of experimental work and method design.ALRP, HT, AP assisted with PCR labworkOR assisted with flow cytometryCL and PD assisted in analysis of resultsDJD, DT, GH supervised, helped with analysis and writing of the manuscriptAll authors reviewed the manuscript

Elson, J., et al., Random intracellular drift explains the clonal expansion of mitochondrial DNA mutations with age. The American Journal of Human Genetics, 2001. 68(3): p. 802-806.
Zheng, W., et al., Origins of human mitochondrial point mutations as DNA polymerase γ-mediated errors. Mutation Research/Fundamental and Molecular Mechanisms of Mutagenesis, 2006. 599(1-2): p. 11-20.
Trifunovic, A., et al., Premature ageing in mice expressing defective mitochondrial DNA polymerase. Nature, 2004. 429(6990): p. 417-23.
Riggs, B.L., et al., Differential changes in bone mineral density of the appendicular and axial skeleton with aging: relationship to spinal osteoporosis. Journal of Clinical Investigation, 1981. 67(2): p. 328.
Riggs, B.L., et al., Changes in bone mineral density of the proximal femur and spine with aging: differences between the postmenopausal and senile osteoporosis syndromes. Journal of Clinical Investigation, 1982. 70(4): p. 716.
Law, M.R., N.J. Wald, and T.W. Meade, Strategies for prevention of osteoporosis and hip fracture. BMJ: British Medical Journal, 1991. 303(6800): p. 453.
Hendrickx, G., E. Boudin, and W. Van Hul, A look behind the scenes: the risk and pathogenesis of primary osteoporosis. Nature Reviews Rheumatology, 2015. 11(8): p. 462-474.
Dobson, P.F., et al., Mitochondrial dysfunction impairs osteogenesis, increases osteoclast activity, and accelerates age related bone loss. Scientific reports, 2020. 10(1): p. 1-14.
Lawless, C., et al., The rise and rise of mitochondrial DNA mutations. Open biology, 2020. 10(5): p. 200061.
Taylor, R.W., et al., Mitochondrial DNA mutations in human colonic crypt stem cells. The Journal of clinical investigation, 2003. 112(9): p. 1351-1360.
Tuppen, H.A.L., et al., Mitochondrial DNA mutations and human disease. Biochimica et Biophysica Acta (BBA)-Bioenergetics, 2010. 1797(2): p. 113-128.
Fayet, G., et al., Ageing muscle: clonal expansions of mitochondrial DNA point mutations and deletions cause focal impairment of mitochondrial function. Neuromuscular Disorders, 2002. 12(5): p. 484-493.
Samuels, D.C., et al., Recurrent tissue-specific mtDNA mutations are common in humans. PLoS genetics, 2013. 9(11): p. e1003929.
Hipps, D., et al., Detecting respiratory chain defects in osteoblasts from osteoarthritic patients using imaging mass cytometry. Bone, 2022. 158: p. 116371.
Chan, C.D., et al., Co‑localisation of intra‑nuclear membrane type‑1 matrix metalloproteinase and hypoxia inducible factor‑2α in osteosarcoma and prostate carcinoma cells. Oncology Letters, 2021. 21(2): p. 1-1.
Drescher, H., S. Weiskirchen, and R. Weiskirchen, Flow Cytometry: A Blessing and a Curse. Biomedicines, 2021. 9(11).
Miltenyi Biotec, MSC Phenotyping Kit human. 2011: http://www.miltenyibiotec.com/en/products-and-services/macs-flow-cytometry/reagents/kits-and-assays/msc-phenotyping-kit-human.aspx.
Dominici, M., et al., Minimal criteria for defining multipotent mesenchymal stromal cells. The International Society for Cellular Therapy position statement. Cytotherapy, 2006. 8(4): p. 315-317.
Coxhead, J., et al., Somatic mtDNA variation is an important component of Parkinson's disease. Neurobiology of aging, 2016. 38: p. 217. e1-217. e6.
Li, H. and R. Durbin, Fast and accurate short read alignment with Burrows–Wheeler transform. bioinformatics, 2009. 25(14): p. 1754-1760.
Li, H., et al., The Sequence Alignment/Map format and SAMtools. Bioinformatics, 2009. 25(16): p. 2078-9.
Li, H., A statistical framework for SNP calling, mutation discovery, association mapping and population genetical parameter estimation from sequencing data. Bioinformatics, 2011. 27(21): p. 2987-93.
Weissensteiner, H., et al., mtDNA-Server: next-generation sequencing data analysis of human mitochondrial DNA in the cloud. Nucleic Acids Res, 2016. 44(W1): p. W64-9.
Goto, H., et al., Dynamics of mitochondrial heteroplasmy in three families investigated via a repeatable re-sequencing study. Genome biology, 2011. 12(6): p. 1-16.
Zerbino, D.R., et al., Ensembl 2018. Nucleic Acids Res, 2018. 46(D1): p. D754-D761.
McElhoe, J.A. and M.M. Holland, Characterization of background noise in MiSeq MPS data when sequencing human mitochondrial DNA from various sample sources and library preparation methods. Mitochondrion, 2020. 52: p. 40-55.
Xu, J., et al., Single nucleotide polymorphisms in the D-loop region of mitochondrial DNA is associated with the kidney survival time in chronic kidney disease patients. Ren Fail, 2015. 37(1): p. 108-12.
Lin, G.L. and K.D. Hankenson, Integration of BMP, Wnt, and notch signaling pathways in osteoblast differentiation. Journal of cellular biochemistry, 2011. 112(12): p. 3491-3501.
Greaves, L.C., et al., Mitochondrial DNA mutations are established in human colonic stem cells, and mutated clones expand by crypt fission. Proceedings of the National Academy of Sciences, 2006. 103(3): p. 714-719.
Mareschi, K., et al., Expansion of mesenchymal stem cells isolated from pediatric and adult donor bone marrow. Journal of cellular biochemistry, 2006. 97(4): p. 744-754.
Greaves, L.C., et al., Clonal expansion of early to mid-life mitochondrial DNA point mutations drives mitochondrial dysfunction during human ageing. PLoS genetics, 2014. 10(9): p. e1004620.
Greaves, L.C., et al., Comparison of mitochondrial mutation spectra in ageing human colonic epithelium and disease: absence of evidence for purifying selection in somatic mitochondrial DNA point mutations. PLoS genetics, 2012. 8(11): p. e1003082.

No competing interests reported.

SupplementarymaterialAppendix1patientdetailstable.docx

Download PDF

Version 1

posted

You are reading this latest preprint version

Variant load of Mitochondrial DNA in single human Mesenchymal stem cells

Status:

Version 1

Abstract

Figures

Introduction

Materials and Methods

Sample Cohort and MSC isolation

Bioinformatic Analysis

Single Cell Quality Control

Statistical analysis

Results

Variant Data

Data Availability

Discussion

Conclusion

Declarations

Author Contribution

References

Additional Declarations

Supplementary Files

Status:

Version 1