Be in it for the long haul: a commentary on human tissue recovery initiatives

Abstract

The strong need for a new foundational molecular framework for human nervous system research at the nociceptive level is now matched by comprehensive and quantitative capabilities for analyzing nociceptive tissues such as pathological peripheral tissue, damaged peripheral nerve, dorsal root ganglia, spinal cord, and brain, where possible. However, this idea must be matched by equally strong organization and infrastructures for multisite tissue recovery, molecular analyses, data sharing, and long-term archiving. Experience from other human tissue analysis projects shows that a decades long activity may be expected, hence “Be in it for the long haul.” While certain milestones can be met fairly quickly, others aimed at molecular and neuroanatomical characterization of chronic pain disorders will require the sustained attention of the groups involved. This can yield a valuable addition to basic and translational pain research and the development of new treatments whose targets are validated directly in humans.

Inclusion of direct-in-human molecular-level investigations of DRG, spinal cord, and potentially brain, into pain research programs is long overdue. Recently, several attempts have been made to articulate a cohesive strategy for enhancing translational research by including transitional animal models, enhancing validation of preclinical targets in human tissue, and coordinating behavioral phenotyping and molecular analyses.^{11, 15, 19, 27} The end goal of these efforts is to address the failure at the “translational bridge” between rodent models and clinical success. The proposed formation of a “Human Cells and Networks of Pain (HCNP) Consortium,” a harmonized, comprehensive infrastructure for patient phenotyping, surgical and post-mortem tissue procurement, as well as sample processing, has the capacity to improve the translational capabilities of the pain research community.¹⁹ Access to phenotype-defined, high-quality human neuronal tissues meets the unique needs for molecular discovery using transcriptomic, proteomic, lipidomic, neuroanatomical, and genomic sequencing methods. Proposals to establish integrated platforms for human tissue research constitute deliberate counterpoints to the long-standing reliance on rodent models, particularly mouse models, for studying nociceptive cellular and molecular networks. In this case, nociceptive circuit is defined as DRG, spinal cord, and damaged peripheral nerves or tissues. This is in contrast to the brain where pain is perceived and/or processed. It is also recognized that inclusion of larger animal models is an essential element to building a stronger foundation for translational validity.¹⁰ How soon can these goals be realized? This depends on good planning and, importantly, consistent long-term funding for the scope of questions being posed. The long-term nature of acquiring adequate numbers of appropriate dorsal root ganglia (DRG), matching spinal cord segments, and even brain regions from donors is a key point that we examine further here. Several early studies have paved the way for direct in human molecular investigations^{5, 13, 16–18, 23, 25} and provide a practical framework for prospective longitudinal acquisitions. One example in the neurosciences is the Alzheimer’s Disease post-mortem brain recovery Religious Orders Study of Bennett and colleagues.¹ It is essential that an active and sustained procurement effort be established that incorporates obtaining tissue from as many pain conditions as possible. Experience shows this will be a multiyear effort. In fact, the title of this commentary reflects this idea and the fact that the longitudinal Religious Orders Study has been operational for nearly 30 years.

To avoid prolixity, many pragmatic points critical to success are typically left out of proposals to study human tissue. For example, there are several known challenges related to PNS and CNS tissue recovery from organ donors or at autopsy.¹³ One example is a recently completed transcriptomic and neuroanatomical study of human DRGs from organ donors with diabetic peripheral neuropathy.⁷ Collecting a set of patient DRGs with the appropriate diagnosis and appropriate somatotopy for their pain condition took nearly two years. A nationwide program might be able to coordinate tissue recovery among organ transplant organizations to overcome bottlenecks in identifying subjects with a particular pathological pain condition for which analysis nociceptive tissues would be informative. To meet this challenge, coordination and harmonization of recovery procedures and donor phenotyping between sites become important issues. On a quantitative basis, the demographics of pain and the frequency of a particular pain disorder in the population are interdependent factors since only 20% of donors may have pain issues based on an analysis of ~32,000 adults across the USA.²⁸ While this estimate may be low, it is important to consider that not all pain diagnoses are homogeneous or precisely phenotyped to the degree necessary for a controlled molecular study.

On a more humanistic level, compassionate communication with next of kin is critical for obtaining consent to recover tissues for research purposes. The consent for tissue often occurs in a chaotic environment for the patient or the family, as end of life care is complex and taxing. Successful consent benefits from an interpersonal connection between the consenter and the family, as well as the ability to create a connection to the larger purpose of the work. Permission is not always forthcoming for every case, which makes acquiring large numbers of patients a slow process. It must be recognized that the emotional strain at the time of donation can be substantial for family members. Thus, staffing the program with experienced, sensitive and dedicated individuals is a critical component and all of the elements for donation can already be in place. Another practical issue at this stage are burial wishes of the family and funeral service planning. The choice of embalming or cremation can be permissive of or limit access to upper levels of the neuraxis such as medullary dorsal horn and tissues that might include the nodose ganglion or brain. One potential path forward is community engagement to reinforce connections with pain patients, families, and even advocacy groups to convey the value of these tissues for development of better solutions for pain control.

Accumulating enough subjects to carefully assess and account for clinical variability is a critical parameter for success and a critical factor for evaluation of underlying genetic susceptibilities. Lessons can be gleaned from the psychiatry field where polygenic risk scores have been derived from up to 100,000 individuals in diseases such as major depression and substance use disorders.^{20, 24} Similarly, pain sensitivity is complex and polygenic, suggesting that finding significant genetic contributions may require very large numbers of carefully phenotyped homogeneous patients. In the context of a tissue recovery program, the N for any one pain indication will not approach the numbers needed to find polygenic risk factors without a large multi-institutional effort.^{3, 14} While some forms of pain can have genetic origins, most pain patients develop their condition after tissue injury, disease pathology, or aging and degeneration over time. The exact nature of the premortem pain etiology can be difficult to fully assess and it may be difficult to obtain thorough documentation of disease and medication status. For prospective studies these considerations reinforce establishing relationships with the individual and their treatment team during the transition to death.

At a more basic level, it might be assumed that the presentation of the pain phenotype (e.g., spontaneous pain) would match the molecular phenotype. Our results suggest this is not always the case. Some diabetic patients in the Hall et al., study had neuropathic pain symptoms but little obvious DRG pathology, either histologically or at the molecular level, whereas obvious signs of cell damage and loss were present in others. This is consistent with other studies showing that radiologic findings are not always predictive of pain in diseases such as osteoarthritis.⁸ Such underlying clinical heterogeneity underscores the importance of identifying robust, molecularly critical points of vulnerability in the pain pathway. Experience shows that the number of subjects and effect size must be large enough to survive this clinical heterogeneity. Subtle molecular alterations may be difficult to detect, even with large sample sizes. In tissue analyses, these considerations demand a carefully curated set of cases and an adequate number of appropriate controls to estimate and account for heterogeneity. However, some alterations that may be highly robust in inbred animal experiments may ultimately fail to recapitulate in clinical samples due to inhomogeneity of the human population.

Maintenance of a public database is an additional important consideration. A large amount of sequence data on DRG, spinal cord and pathological peripheral tissue is already available for examination of genes that might be potential translational targets for analgesia.^{5, 6, 11, 13, 16–18, 21, 23, 25}. The results from many RNA-seq studies are available through repositories such as dbGaP, the Sequence Read Archive of the NCBI but these contain relatively raw sequence data and need to be processed and placed in a conveniently accessible and dedicated pain website (see commentary by Denk 2017). An example of such a site is Genotype-Tissue Expression Project (GTEx, see Table 1). Enhancing accessibility to pain-specific data can enable future studies going forward.^{13, 17, 23} Our lab and others have contributed sequencing runs from mouse, rat, canine, and human DRG and spinal cord, rat sciatic nerve and, in some cases, with and without nociceptive or other manipulations.²² In the aggregate, these form a growing and comprehensive animal and human “nociceptome.” This can be extended to evaluation of the impact of pain disorders on higher CNS regions that comprise the pain connectome.^{9, 12, 26} One important use of such a database is that it can be consulted to make informed judgements for projects that are being proposed. For target validation, the most fundamental question to ask is “Is the gene expressed in the relevant tissues of the nociceptive circuit (e.g., DRG or spinal cord) in humans?” and, more specifically, “Is it in the cell type of interest?” The very next question is: “To what degree (how much) is it expressed to make it worthwhile for investigation?” A straightforward answer to the latter question is not always possible, especially if the level of gene expression or encoded protein is low.

Table 1.

The table lists various organizations involved in neural tissue banks. Links 1 and 2 are to sources for tissue acquisition that can include spinal cord and DRGs. Link 3 primarily concerns tissue from spinal cord injury patients. Links 4 through 7 are to brain banks that specialize in neurosurgical and neuro-oncological specimens or brains from individuals with neurological, psychiatric, and substance abuse disorders. Links 8 and 9 are to government and accreditation programs for brain banks. Link 10 is to the Genotype-Tissue Expression Project (GTEx) of the National Human Genome Research Institute. Version 8 contains genotype data on 858 donors and RNA-Seq from over 50 peripheral organs and tissues including 12 brain regions and spinal cord. This gives an idea, in part, of what a database might contain, look like, and how it might be used.

	Web site	Comments
1	https://anabios.com/	AnaBios recovers nervous system tissues from organ donors such as spinal cord and dorsal root ganglia for physiological, anatomical, proteomic, metabolomic and gene expression analyses.
2	https://ndriresource.org/	The National Disease Research Interchange’s Human Tissue and Organs for Research Resource (HTORR) provides multiple types of human biospecimens to academic researchers.
3	https://www.themiamiproject.org/	The Miami Project maintains a spinal cord tissue bank from patients that had spinal cord injury.
4	https://www.mcleanhospital.org/research/brain-bank	The Harvard Brain Tissue Resource Center (HBTRC) at McLean Hospital collects and distributes human brain specimens for research.
5	https://www.nimh.nih.gov/research/research-conducted-at-nimh/research-areas/research-support-services/hbcc	The NIMH Human Brain Collection Core (HBCC) recovers brains post-mortem from individuals with neurological, psychiatric and substance use disorders
6	https://medschool.cuanschutz.edu/neurosurgery/research-and-innovation/services/nervous-system-biorepository	The Neural Tissue Bank contains nervous system tissues obtained during surgery such as tumors, brain tissue, dura, and other disease specimens
7	https://www.feinberg.northwestern.edu/sites/nstb/.	The Nervous System Tumor Bank contains patient derived tissues, cells, blood, and CSF mainly from neuro-oncology cases.
8	https://neurobiobank.nih.gov/	The NIH NeuroBioBank (NBB) program integrates policies and procedures of the Brain and Tissue Repositories (BTRs) network for neurological, developmental, and psychiatric disorders research.
9	https://www.aatb.org/	The American Association of Tissue Banks (AATB) is a non-profit, scientific, and educational professional organization for standard-setting and accreditation of over 120 tissue banks.
10	https://gtexportal.org/home/	GTEx is a National Center for Biotechnology Information searchable database containing RNA-Seq data from tissues, organs, and brain regions from over 800 postmortem donors

For all the difficulties that direct-in-human studies pose, their biggest advantage is that they do not need to be translated. However, to ascertain if the results obtained from one cohort are generally applicable to human subjects requires determinations in multiple independent cohorts. Once an apparent and consistent molecular change has been discovered, further work establishing a causal relationship between molecular alterations and the pain problem may be needed. Rare conditions or rare mutations may not be present in many people but can point to fundamental nociceptive mechanisms. Alternatively, some findings may be broadly applicable, and capable of being reproduced in larger cohorts and/or additional pain indications. As a guide from another field, some of the elements and barriers related to a large-scale collaboration were recently discussed in an account of the QBI Coronavirus Research Group collaborative effort.⁴ This is an avalanche that is only just beginning, and as more and more studies focus on the human, the momentum will continue to grow. There is a lot of energy, and in some ways a lot of hopeful naiveté. The process for determining prioritization of pain indications will need to be carefully considered to harness this energy, and the administrative and steering structures will play a critical role.

Optimally, the path forward involves a continuous bidirectional effort with human observational studies brought into preclinical animal models to examine mechanism and causation, and, in the opposite direction, preclinical findings being validated in humans or human tissues. With modern mouse genetics the level of complexity of the questions that can be investigated is far higher than what can be done in human postmortem tissue or patient-derived cell lines. Causation can be very hard to establish using only human tissue. Conversely, relevance to human physiology or pathophysiology can be very hard to assess using only animal models. While we would not expect the closure of mouse facilities and a tectonic shift in translational research, we do think that some questions are easier to answer in human tissue first, before working up a detailed mechanism in preclinical studies. Even at the most fundamental level, knowing that the molecular targets actually are expressed in the human in the same cells and tissues as preclinical rodent models is a prerequisite for a rodent study to be truly translational.

In closing, advancing towards direct-in-human studies is an important step in providing needed foundational data in the pain field and the search for better analgesics. To reiterate the value proposition: These studies indirectly or directly validate animal research by exploring human pathological pain states to obtain a solid molecular human foundation. These insights can only come from the study of donated human CNS and PNS tissues, and ultimately will enhance discovery in both human and animal studies by focusing research on conserved mechanisms, receptors, and biomarkers relevant to human health and disease. The path forward likely contains many pitfalls, barriers, and unexpected obstacles, but also new grains of wisdom that cannot be gained from animal studies, yet can be amplified by concurrent animal or cellular experiments. If human tissue investigations are worthwhile, and we would argue that they are worthwhile, then some of these logistical challenges need to receive more, and sustained, support in order to ensure continuity, rigor, reproducibility, and full participation of the pain research communities, patients, and their families. Given the anticipated long duration of this endeavor, sustaining the momentum of the community is an important factor.

Perspective:

A concerted effort is needed to build human nociceptive tissue banks for multi-omic research. In addition to collecting tissue, a careful characterization of pain problems from donors is essential, as is a parallel effort to assess their concurrent medical problems, medications, and the many variables of general human activity and lifestyle that can impact the results. Given the projected long time frame, in addition to maintaining funding, sustaining motivation and momentum are critical factors for success.

Highlights.

• Translational efforts can be facilitated by investigations of human nervous system tissues

• Putative preclinical analgesic targets must be expressed in the relevant human cells

• Establishing human relevance enhances the value of mechanistic animal studies

• Multi-omic databases of human tissue data are growing rapidly

• Human tissue investigations require sustained research support and planning