Dr. Mariya Mardamshina, a postdoctoral fellow at Stanford University, is pioneering spatial proteomics technologies that reveal cellular states previously invisible to researchers. Her journey from medical doctor to spatial proteomics researcher has taught her that biological systems rarely fit into neat categories. It’s this embrace of cellular complexity, particularly the transitional states that challenge textbook definitions, that’s reshaping our understanding of cellular identity and disease, and driving her toward discoveries with potential therapeutic impact.

From discovery engine to precision tool

Initially trained as a medical doctor, Mariya found herself wondering what lies beyond what we can see on a standard histology slide. This curiosity led her to the molecular language that cells use, particularly proteins, which are “the true workhorses of biology” to Mariya. With proteomics, she discovered tools that could explore not just what a cell is, but where it is, what it’s doing, and how it changes in context. “I was fascinated by how proteomics can serve as a functional readout of cellular state,” she explains. “Early in my training, I saw a gap between transcript-level information and the actual phenotype. Proteomics offered a more direct route to understanding cell function. In the context of disease, that’s critical.”

Mariya’s research has evolved from broad discovery proteomics to spatially resolved approaches that preserve native tissue architecture. Her recent breakthrough with mxDVP (Multiplexed Deep Visual Proteomics) exemplifies this evolution. The team achieved deep profiling of over 6,000 proteins from just 100 spatially selected cells within intact human pancreatic islets and were able to detect rare polyhormonal endocrine states that conventional methods would have missed, marking the first successful application of spatial single-cell proteomics to intact islets. This milestone required tremendous technical refinement, collaboration, persistence, and a deep respect for the biological complexity of the system.

“Islets themselves are incredibly challenging: they’re small, densely packed, and contain a mosaic of rare endocrine states that are difficult to resolve,” Mariya notes.  “Plus, islet cells are four times smaller than those typically analysed in proteomics workflows.” The achievement was enabled by meticulous optimisation of single-cell segmentation, precise contour-aligned laser dissection, ultra-sensitive mass spectrometry, and reliable analytical tools like IonOpticks columns. “Their enhanced reproducibility and robustness have been essential for our spatial proteomics workflows,” Mariya notes. “The improved peptide separation and sensitivity is essential when working with low-input samples”.

The results were remarkable: the identification of 12 distinct endocrine subtypes, including rare polyhormonal states that challenge classical definitions of pancreatic cell types. “These findings open the door to understanding cellular plasticity in metabolic disease and regeneration. To me, this work exemplifies a broader transformation in the field: proteomics is no longer just a discovery engine; it’s becoming a precision tool, capable of dissecting cellular ecosystems with near-surgical resolution.”

With the continued advancement of these technologies, Mariya sees “tremendous promise for clinical translation, from mapping disease heterogeneity to guiding targeted interventions for previously untreatable diseases,” as exemplified by a recent Schweizer et al. study and Nordmann et al.’s 2024 study, respectively.

Learn more about mxDVP in Mariya’s latest manuscript.

Embracing the unexpected

One of the most transformative moments in Mariya’s research came during the pancreatic islet project, when her team began detecting cells that co-expressed markers from multiple endocrine lineages. Initially, these rare polyhormonal states seemed like potential artifacts – segmentation errors or staining bleed-through that would be easy to dismiss.

“But their spatial consistency, reproducibility across donors, and distinct proteomic signatures compelled us to look deeper,” she recalls. Validating these findings made it clear that these weren’t noise; they were real, spatially organised cell populations that challenged long-standing assumptions about islet composition. Many showed signs of functional plasticity, suggesting they might represent transitional states involved in metabolic adaptation or regeneration.

“This discovery changed how I think about cellular identity,” Mariya reflects. “We often aim to categorise cells neatly into defined types, but biology doesn’t always work that way. These transitional or plastic states may be rare, but they could hold critical insight into how islets maintain resilience under metabolic stress” – a finding that could ultimately inform therapeutic strategies aimed at preserving or restoring endocrine function.

Beyond the biological implications, the discovery underlined the value of their technical approach. “It highlighted the power of combining refined spatial stratification with deep proteomic profiling, allowing us to access biological states that conventional marker panels would likely miss,” Mariya explains. This experience reinforced for Mariya the importance of not excluding unexpected data points. “Sometimes, the most informative signals come from the cells that defy classification,” she notes.

Lessons from tumour heterogeneity

The lessons about embracing unexpected findings trace back to her PhD work on breast cancer heterogeneity, a project that stands out as one of the most formative experiences in her career. Profiling over 300 regions across 35 tumours using spatial proteomics, her team discovered that even tumours appearing homogeneous by conventional pathology revealed deep, structured proteomic heterogeneity.

Some of the most revealing tumour regions looked “bland” under the microscope – nothing a pathologist would flag – but proteomically, they were enriched for immune-evasive or stem-like programs. “Initially, we considered discarding a few of these samples as ‘uninformative’, but when we looked more closely, they turned out to hold unique insights,” she remembers.

That was a turning point. “I realised how powerful spatial proteomics can be, not just in capturing what’s there, but in revealing hidden subpopulations that influence tumour behaviour and treatment response,” Mariya reflects. This reshaped her understanding of tumour classification and personalised therapy, revealing tumours as evolving ecosystems rather than static entities. “If we want to treat them effectively, we need tools that capture that complexity,” Mariya adds.

Navigating technical challenges

The development of mxDVP wasn’t without its obstacles. The biggest challenge was achieving precision at every step: imaging, segmentation, dissection, and mass spectrometry, without compromising throughput or data quality. With such low input amounts, there was no margin for error.

“Even a single misaligned cell or imprecise cut could jeopardise the entire sample,” Mariya explains. “That created a constant sense of pressure; every run truly counted.” There were moments Mariya described as feeling “like we were walking a tightrope”. She still remembers the collective tension the first time the team ran samples through the full pipeline: “we all knew how much had to go exactly right for it to work. But we didn’t back down.”

The team developed a dedicated image analysis suite (PIPΣX), integrated it with automated laser microdissection, and iterated obsessively. Achieving success in something that “once felt just out of reach” required not just technical problem-solving but emotional investment from a multidisciplinary team with shared ownership and purpose. “The strength of the team is what made the impossible feel doable.”

Perhaps even more rewarding was the idea that the flexibility and adaptability of the workflow they built can empower others. “That feeling is hard to put into words,” Mariya reflects.

Looking back and looking ahead

As Mariya reflects on her journey, she offers advice to her younger self: “I’d tell myself to embrace the unknown and to stop apologising for asking too many questions. Some of my most important projects started as naive ‘what-ifs.'” She also wishes she had learned earlier that imposter syndrome is common even among brilliant people, and that confidence comes from doing hard things repeatedly while staying curious through discomfort.

The future of proteomics excites her immensely. “The convergence of high-plex imaging, spatial transcriptomics, and ultra-sensitive proteomics is electrifying,” she says. “Tools, including IonOpticks columns, have made it possible to ask and answer questions that seemed unreachable just a few years ago. With tools like mxDVP, we can now ask: What proteins are present, where, and in whom, all from a single FFPE slide.” This capability will fundamentally shift how researchers study disease, enabling the construction of spatial atlases of pathogenesis, regeneration, and treatment response.

The field is also becoming more democratised, with better software and sample preparation protocols bringing us into an era where spatial proteomics might become as routine as immunohistochemistry. For Mariya, this democratisation represents hope for broader impact, particularly in understanding and treating metabolic diseases where her islet work might one day contribute to restoring pancreatic function.

“I’ve seen firsthand how the loss of pancreatic function can profoundly affect someone’s life,” she reflects. “Knowing that our work might one day contribute to restoring this complex organ’s function brings an added layer of meaning to what we do in the lab each day.”

Her gratitude extends to the global proteomics community, a field she describes as uniquely generous and collaborative, where sharing knowledge and building on each other’s innovations is the norm. “It’s the people behind the tools, the collaborations, conversations, and shared discoveries that make this work truly meaningful,” she concludes. “I feel incredibly lucky to be part of one of the most supportive and inspiring scientific communities I’ve known.”

Would you like to be featured in an upcoming Community Newsletter? Email us at [email protected] and tell us a bit about yourself and your research.