Tutor Lesson

The body contains many kinds of cells arranged into tissues, organs, and organ systems. These cells vary in shape and purpose even though almost every cell carries essentially the same DNA. The key question is how a shared genome can produce such different cells. The Human Cell Atlas approaches this problem by determining which genes are active in each cell, where the cell lies, and how it relates to nearby cells.

Cells do not activate every gene in the same pattern. Regulatory machinery leaves some genes inactive and switches others on. A fibroblast uses genes associated with connective tissue, whereas a neuron uses genes needed for electrical signaling and communication. The resulting proteins give these cells different structures and roles.

By about 2013, genomic technology could examine the tiny amounts of nucleic acid contained in one cell. This “resolution revolution” shifted research away from mixtures of cells and toward cell-by-cell measurement.

Bulk genomics analyzes many cells together, so its result averages their measurements. In a mixed tissue, a signal from a rare population can disappear beneath signals from abundant cells. Single-cell approaches measure cells separately and preserve distinctions that an average can conceal.

• Determine which genes are active in each cell.
• Recognize the cell types contained in a tissue.
• Compare molecular differences among cell populations.
• Detect rare or previously unrecognized cell types.
• Investigate how cells are positioned and how they interact in tissue.

The Human Cell Atlas is an international effort to chart the types and properties of human cells. Its scientific aim is to explain bodily organization at cellular resolution. Its medical aim is to support the study of health, disease, diagnosis, monitoring, treatment, toxicity, rare variants, and tissue repair. Because disease is recognized through change, a detailed healthy reference is needed before a cell population, expression pattern, or tissue arrangement can be judged abnormal.

• The cell types present in every tissue.
• The molecular features of those cells.
• Their positions inside tissues.
• The neighboring cells with which they interact.
• Population changes during development, aging, or disease.
• Signals that shape cellular behavior.

Tissue- and organ-focused biological networks are joined by groups concerned with computational methods, standards, technology, metadata, development, organoids, genetic diversity, equity, and diversity. Sequencing alone cannot make an atlas: samples require consistent collection, accurate analysis, and metadata linking every cell to its tissue source. With roughly 50 tissues arranged into 14 organ systems, the body must be assembled through many coordinated studies.

At the time presented, the data platform held millions of cells from many people and samples, including kidney, skin, lung, gut, liver, and embryonic and fetal tissues. The atlas was therefore accumulating organ by organ and tissue by tissue.

Single-cell genomics must prevent RNA from one cell being mistaken for RNA from another. In one microfluidic approach, cells enter separate chip compartments, are lysed so their membranes open, and release their RNA. A cell-specific molecular barcode is attached to that RNA, allowing later sequencing and analysis to trace each molecule to its original cell.

• Microfluidic droplets enclose individual cells in tiny drops.
• Nanowell plates separate cells into small wells.
• Multiplex methods apply identifying barcodes.
• Split-sequencing, or split-seq, uses a related labeling strategy.

1. Capture or separate individual cells.
2. Lyse each cell and release its RNA.
3. Amplify the scarce RNA material.
4. Build molecular libraries for sequencing.
5. Sequence the RNA molecules that are present.
6. Computationally assign transcripts to cells and identify populations.

The final ambition extends beyond a catalogue. It is a three-dimensional map of cell locations, exchanged signals, and local neighborhoods or niches. In this view, tissue function emerges from interacting cellular circuits as well as from each cell's individual properties.

Single-cell RNA sequencing supplies detailed expression profiles but commonly removes cells from their original positions. Spatial transcriptomics preserves location by placing a tissue section on a chip whose capture features collect RNA from the regions above them, connecting detected expression to coordinates in the tissue.

A capture feature can be about 50 micrometers wide and may contain RNA from roughly 20–40 cells, depending on tissue architecture. The resulting map is broad and comprehensive but is not always a true single-cell measurement.

• Single-cell RNA sequencing supplies detailed expression signatures for separate cells.
• Spatial transcriptomics maps expression broadly and relatively without bias across sections.
• Multiplex FISH detects many RNA markers and locates individual cells precisely.
• Computational analysis joins the measurements into one consistent map.

Together these techniques approach a “Google Maps” representation that combines large-scale tissue structure with fine-grained cellular identity.

The uterus is difficult to map because it changes greatly through the menstrual cycle. Its inner lining, the endometrium, is shed during menstruation, rebuilt by dividing cells in the proliferative phase, and transformed during the secretory phase as epithelial cells release nutrients and prepare the tissue for implantation. The surrounding myometrium is smooth muscle.

• The basal layer lies deepest and is associated with stem cells, although their exact identity and location remain unresolved.
• The glandular layer contains the uterine glands.
• The luminal layer faces the uterine interior.

• The tissue changes throughout each monthly cycle.
• Structures such as spiral arteries have features especially associated with primates.
• Human uterine material is hard to obtain.
• Sampling methods capture different depths and cycle stages.

Researchers studied diagnostic biopsies later judged healthy and tissue from deceased transplant donors. Biopsies usually sampled the upper endometrium and provided denser coverage of cycle stages, whereas rare donor tissue could preserve the uterus's full depth. Single-cell and single-nucleus RNA sequencing, spatial transcriptomics, and imaging then identified populations and positioned them within the layers.

Endometrial organoids can originate from adult stem cells in tissue samples. Researchers digest the tissue, place it in a supportive material such as Matrigel, grow and passage it for several weeks, and then compare single-cell RNA-sequencing profiles from the organoids with cells in the living uterus.

1. The in vivo atlas helps researchers refine laboratory-grown tissue.
2. The in vitro organoids provide a system for testing mechanisms that regulate cell behavior.

• Cell-surface receptors.
• Secreted signaling factors.
• Ligands that bind those receptors.
• Factors to place in the culture medium.
• Extracellular matrices or scaffolds that support cells in relevant tissues.
• Signals that direct particular maturation routes.

Changing signaling conditions altered endometrial organoid fate. One condition yielded twisted organoids dominated by ciliated cells. Manipulating Notch signaling instead drove a secretory lineage, producing more columnar cells resembling glandular epithelium. In this sense, atlas data works like a recipe book for targeted tissue engineering.

After implantation, the endometrium becomes the decidua. Maternal natural killer cell subsets lie near fetal extravillous trophoblasts, which extend outside the embryo and carry paternal genetic material. This poses an immune challenge: paternal alleles are not identical to maternal alleles, yet the fetus must be tolerated rather than rejected.

Atlas measurements exposed immunosuppressive and immune-modulating signaling involving innate and adaptive processes. The interface is therefore not immunologically silent. It remains active but controlled, permitting pregnancy while retaining appropriate regulation.

Natural killer cell–trophoblast interactions have been associated with incompatibilities between maternal and paternal alleles that can contribute to pre-eclampsia and miscarriage. Directly charting human tissue is especially useful because this cellular setting can differ markedly from laboratory-animal systems.

• Bronchi form the larger airway passages.
• Smaller airways and alveoli occupy deeper lung regions.
• Alveolar cells contribute to gas-exchange surfaces.
• Goblet and ciliated cells support airway epithelial function.

Asthmatic inflammation contributes to thickening of airway muscle and can culminate in an attack. Researchers compared healthy references with bronchoscopy material from living people with asthma and lung parenchyma from deceased transplant donors. They found an abnormal bronchial epithelial pattern and distinctive T-cell populations. Those T cells may offer targets for new treatments, particularly when steroid therapy is inadequate.

Deep lung tissue contains type 1 and type 2 alveolar cells; larger airways contain populations including goblet and ciliated epithelial cells. Goblet and ciliated cells also occur in the nose, but a shared broad label does not make cells molecularly equivalent. Likewise, uterine secretory cells and airway goblet cells both release material yet have gene-expression programs and functions adapted to different tissues.

Before the COVID-19 pandemic, airway research had found a nasal goblet-cell population with comparatively high ACE2 expression. Because SARS-CoV-2 uses ACE2 for cellular entry, researchers could rapidly reuse existing and shared atlas data to ask which cells across the body express this receptor and related entry machinery. This shows how a reference collected for one purpose can become useful during an unexpected emergency.

Nasal expression may help explain early airway interaction and efficient transmission. Type 2 alveolar expression had previously been detected with antibodies and immunohistochemistry. The conjunctiva also expressed a protease involved in preparing the viral surface for entry, and its connection to the nose through the nasolacrimal system suggested a possible ocular–nasal route. Ileal expression showed that relevant receptors were not restricted to the respiratory tract. Placental expression offered clues for studying rare maternal-to-fetal transmission.

No single-cell RNA-sequencing protocol recovers a perfectly complete transcriptome from every cell. Full-length methods such as Smart-seq-based approaches can capture more complete transcripts and, at sufficient depth, detect roughly one to ten messenger RNA molecules per cell. Their limitation is lower throughput. Droplet microfluidic methods examine far more cells but usually with less sensitivity.

• Sensitive protocols yield deeper information about each cell.
• High-throughput protocols sample many more cells.

The research question decides which advantage matters. Detecting a very weakly expressed transcription factor favors sensitivity; discovering rare populations across a tissue favors breadth.

1. Survey many cells with a high-throughput method to discover candidate rare populations.
2. Enrich the population of interest.
3. Apply a sensitive full-length protocol for deeper characterization.

Alternative splicing creates multiple transcript isoforms from many genes. Resolving them generally requires full-length RNA capture or sufficiently detailed sequence information, so isoform analysis inherits the same throughput trade-off.

Spatial features can mix neighboring cells. Deconvolution with single-cell references produces quasi-single-cell estimates, not direct boundaries. Microscopy gives finer spatial resolution, while spatial transcriptomics covers expressed genes more comprehensively; their combination is stronger than either method alone.

Single-cell methods detect transcripts encoding mitochondrial proteins from nuclear DNA as well as transcripts originating in mitochondrial DNA. A large mitochondrial-RNA fraction can indicate a damaged cell with a permeable membrane: cytoplasmic RNA leaks away while mitochondrial RNA stays comparatively protected and becomes overrepresented. Mitochondrial reads can therefore carry biological information and also warn of poor sample quality.

RNA abundance is not the same as protein abundance. Translation, protein stability, modification, and degradation are controlled separately. Proteomics could therefore complement transcriptomics, but unbiased protein measurement generally needs far more input—often thousands of cells rather than one. One proposed path is to use atlas information to identify and purify a population, collect enough cells, and then measure its proteins.

Immune cells, vascular cells, and fibroblasts can retain broad similarities across tissues while adapting locally. Epithelial cells are often much more specialized to their tissue. Consequently, labels such as “ciliated” or “secretory” may cover cells with very different molecular identities and functions.

A future collection of cell atlases across species—and eventually across the tree of life—could help reconstruct the evolution and diversification of cell types. Increasing the number of organismal atlases would make those comparisons more accurate.

1. Identity: determine which cell types are present.
2. Gene activity: measure the genes switched on in each cell.
3. Location: establish where each cell resides.
4. Neighborhood: identify nearby cells.
5. Communication: infer the signals cells exchange.
6. Change through time: follow development, menstrual cycling, pregnancy, and disease.
7. Clinical relevance: connect populations and pathways to diagnosis, drug discovery, toxicity, and repair.

The uterus connects cellular identity with layer and menstrual phase. The maternal–fetal interface shows controlled communication between maternal and fetal cells. The asthma study demonstrates disease-related changes in epithelial and immune populations. The COVID-19 work illustrates rapid reuse of a reference to locate potential viral-entry sites. Organoid experiments convert healthy-tissue signals into practical culture instructions.