ProteinSimple Instruments Ella, Maurice, and Jess

Need Protein Validation of Your Single-cell RNA Sequencing (RNA-seq) Data?

Do you have a list of targets from your Single-cell RNA sequencing (RNA-seq) experiments and want to verify protein expression levels for your targets? Only Single-cell Westerns offer the versatility to validate diverse protein targets discovered in your sequencing runs.

View Transcript

Learn About Single-cell Westerns on Milo and Single-Cell RNA Data

With Milo from ProteinSimple, you can perform Single-cell Westerns and measure protein expression in thousands of individual cells in a single run. Good for validating single-cell RNA data at the protein level.

Since mRNA levels do not always correlate with functional protein levels, pairing single-cell RNA data with single-cell protein expression data gives you more accurate and complete conclusions about cellular function.

In this sample, all cells are positive for HIF-1α RNA, but Milo shows that only about half of the cells contain HIF-1α protein.

Milo users are uncovering novel insights at the single-cell level that are leading the publication in the top scientific journals.

How Protein Validation with Single-Cell RNA-Seq Data Works

Here's how protein validation with single-cell RNA-seq data works.

You put your cell suspension on the scWest chip and put the chip in Milo. Milo captures over a thousand cells, lyses them, runs an SDS-PAGE separation on every single-cell lysate, and immobilizes all your separated proteins in just five minutes.

Next, you prep your protein targets on chip with standard primary and fluorescent secondary antibodies and then image chip fluorescence.

Data analysis is then performed for you automatically using ProteinSimple’s Scout software. The entire process from sample loading to data analysis takes just four to six hours with no overnight transfer step.

You can use off-the-shelf primary Western antibodies, so you can be sure you'll find an antibody against all your targets of interest.

Hear From Scientists Using Milo for Gut-Brain Axis Research

Here's what Dr. Maya Kaelberer, a postdoctoral fellow at Duke University, had to say about her experience using Milo to uncover new insights she published in Science about how the gut and brain communicate. Learn more in the Guts and Glory: Validating a Neuroepithelial Circuit using Milo from your peers story.

“In our lab, the way that we've been using Milo is to pair it with kind of single-cell sequencing data so we have the RNA. We know the transcripts are there, but this is actually a way for us to measure and quantify the amount of protein that's in each individual cell. My favorite feature of Milo is definitely how easy it is to use. You basically just put your samples on their slides and pop it into Milo. You run the Gel through for a short period of time, right it's just seconds and having that automated as opposed to manually trying to turn on and off things has made it a lot easier. The power of it, which I really like, is the fact that you can assay a number of cells, in these kind of microwells, and then get their protein and output. So, the software was very easy to use. It’s fairly intuitive which is pretty much what you would want from your software, and it tells you exactly what you need to know. I would recommend Milo to my colleagues, and I have actually discussed it with some of them. We're moving forward with working with it on our NASH study.”

To request quote and learn more about Milo, visit the Milo instrument page.


Milo provides single-cell protein expression information to validate your single-cell RNA data. Since mRNA levels do not always correlate with functional protein levels, pairing single-cell RNA data with single-cell protein expression data is critical to making accurate and complete conclusions about cellular function. Milo uses the large Western catalog of antibodies & can easily measure proteins irrespective of their location in or on a cell, making Milo the only platform with the versatility to detect diverse protein targets that are discovered in your RNA-sequencing runs.

Reguest Pricing

Doing Single-cell RNA-seq?

check Validate RNA heterogeneity data
check Quantify protein heterogeneity
check Confirm gene expression results
check Correlate RNA-seq data to protein expression on a single-cell level


The following RNA-seq data validation information is presented in collaboration with Dhananjay Wagh, Ph.D. and John Coller, Ph.D., Stanford Functional Genomics Facility, Stanford University School of Medicine, Palo Alto, CA and Jason Stafford, Ph.D. and Kelly Gardner, Ph.D., ProteinSimple, San Jose, CA.

Workflow for protein validation of single-cell RNA-seq data Figure 1. Workflow for protein validation of single-cell RNA-seq data.


The Milo Single-Cell Western platform provides scientists with protein validation data for their single-cell RNA-seq gene expression data (Figure 1). Milo users can measure protein expression levels in over 1,000 individual cells per run and can multiplex with typical assays detecting approximately four proteins per cell simultaneously using a variety of multiplexing approaches.

Regardless of what targets are uncovered in your RNA-seq run, Milo has the flexibility and versatility to detect them at the protein level. Milo uses commercially available western blot antibodies, giving users access to the broadest set of detection reagents to validate even uncommon targets that emerge from their sequencing runs. Furthermore, users can measure proteins irrespective of where they are located in or on the cell, measuring transcription factors and protein isoforms which can be challenging to measure by flow cytometry. Single-cell Westerns can also be used to study post-translational modifications such as phosphorylation that are not revealed by RNA-seq analysis.

HIF-1α protein regulation HIF-1α protein is rapidly degraded when O2 is available (normoxia)
Figure 2. HIF-1α protein regulation. HIF-1α protein is rapidly degraded when O2 is available (normoxia). When O2 is not available (hypoxia) HIF-1α escapes degradation and acts as a transcription factor.


Proteins can be rapidly degraded after they are translated, leading to significant differences in the levels of mRNA versus functional protein within a cell. Expression of the transcription factor known as Hypoxia inducible factor 1-alpha (HIF-1α) is a classic example. HIF-1α mRNA and protein are constitutively expressed in cells. However, under normoxic conditions (when O2 is readily available) HIF-1α is rapidly ubiquitinated, targeting it to the proteasome for degradation (Figure 2). This process is mediated by an oxygen-dependent prolyl hydroxylase (PHD) and an E3 ubiquitin ligase known as von Hippel-Lindau (VHL) protein.

Under hypoxic conditions (when O2 is scarce), PHD activity is inhibited and HIF-1α escapes proteosomal degradation. HIF-1α heterodimerizes with HIF-1β to form a transcriptionally-active complex that regulates the expression of >60 genes including vascular endothelial growth factor (VEGF) and erythropoietin (EPO), signaling molecules important for increasing O2 delivery to hypoxic tissues. As a result, HIF-1α mRNA and protein levels only correlate under hypoxic conditions but differ substantially under normoxic conditions.

Single-cell RNA-seq analysis of HIF-1α
Figure 3. Single-cell RNA-seq analysis of HIF-1α. A mixture of DFO-treated and untreated HeLa cells was assayed for expression of HIF-1α. 100% of the cells were found to express HIF-1α RNA. Cell clustering based on β-tubulin expression revealed a single, homogeneous population.


Highlighting the need to validate gene expression at the protein level, Milo was used in parallel with single-cell RNA-seq experiments to validate HIF-1α protein levels in a heterogeneous cell population consisting of cells subjected to either hypoxic or normoxic conditions.

HeLa cells were treated for 24 hours with 0.5 mM deferoxamine (DFO), an iron-chelating agent that mimics hypoxia by inhibiting PHD activity. Untreated HeLa cells (normoxic conditions) were used as a control. Manufacturer’s guidelines were followed for cell preparation. Both treated and untreated cells were dissociated, washed, counted, and mixed at a 1:1 proportion.

The mixed sample was then split and analyzed in parallel on the Milo Single-Cell Western system and the Chromium Controller system using the v2 Chromium Single-Cell 3’ Solution followed by sequencing on a NextSeq 500 (Illumina) (workflow shown in Figure 1).


One thousand cells (DFO-treated and untreated mix) were targeted for capture using the 10x Chromium Controller and Chromium Single Cell 3' v2 chemistry. The manufacturer’s protocol was followed for cDNA amplification, fragmentation, and library preparation without any modifications. The library was sequenced on a NextSeq 500 instrument. Data was analyzed using cell ranger (version 1.3.1). A total of 204 cells were captured. About 1.5 million reads per cell were obtained. Normalized expression values for HIF-1α and β-tubulin (BTUB) were extracted from each cell and a scatter plot of HIF-1α vs BTUB was created (Figure 3).


One thousand cells (DFO-treated and untreated mix) were targeted for capture on an scWest chip using the standard Single-cell Western workflow. Briefly, 1 mL of a 100,000 cell/mL cell suspension was loaded onto a Large scWest chip and allowed to settle for 10 minutes.

Single-cell occupancy was confirmed via brightfield microscopy and the scWest chip was then run on the Milo instrument with the following conditions: 10 sec lysis, 75 sec electrophoresis at 240 V, and 240 sec UV exposure. The chip was then simultaneously probed for 2 hours with a primary antibody cocktail of mouse anti-HIF-1α antibody at 100 µg/mL final concentration and rabbit anti-β-tubulin antibody at a 50 µg/mL final dilution.

After washing 3x15 min with Wash Buffer, the chip was then probed for 1 hour in the dark with a secondary antibody cocktail containing donkey anti-mouse IgG Alexa Fluor 647 and donkey anti-rabbit IgG Alexa Fluor 488, each at a final concentration of 100 µg/mL. The chip was then washed and imaged on an Axon 4400a microarray scanner (Molecular Devices). Images were analyzed using Scout Software (ProteinSimple) to quantify peak areas for HIF-1α and β-tubulin in each single-cell lysate.

Single-Cell Western analysis of HIF-1α where HIF-1α protein was detected in only a subset of HeLa cells as identified by β-tubulin expression

Single-Cell Western analysis of HIF-1α where Bivariate analysis revealed that ~46% of analyzed β-tubulin+ cells were HIF-1α+

Figure 4. Single-cell Western analysis of HIF-1α. A) HIF-1α protein was detected in only a subset of HeLa cells as identified by β-tubulin expression. Example separation images of six cells are shown with HIF-1α protein visualized in the 647 channel and β-tubulin visualized in the 488 channel. Green electrophoresis lanes indicate lanes where peaks were identified by Scout Software, whereas blue lanes indicate lanes with no peaks. Fluorescence intensity plots generated by Scout Software are shown on the right for all β-tubulin+ lanes. B) Bivariate analysis revealed that ~46% of analyzed β-tubulin+ cells were HIF-1α+.

RNA-seq analysis revealed two distinct cell populations

Figure 5. RNA-seq analysis revealed two distinct cell populations. A heat map of gene expression shows that the mixture of DFO-treated and untreated HeLa cells contained two populations with distinct gene signatures. Both populations were positive for HIF-1α mRNA. Milo can be used to validate protein expression for these target candidates to understand their functional role in cellular function.

Cellular Subpopulations Identified by Single-Cell Westerns

RNA-seq analysis of the HeLa cell mixture showed that 100% of the cells expressed HIF-1α mRNA (Figure 3). However, Single-cell Western analysis with Milo revealed that only 46% of the cells identified by β-tubulin expression stained positive for HIF-1α protein (Figure 4). As expected, HIF-1α protein expression was degraded in the population of cells exposed to normoxic conditions. Single-cell protein expression did not correlate with single-cell RNA expression, highlighting the need for both protein and RNA expression information in single-cell gene expression studies.

The single-cell RNA-seq data also identified two distinct cell populations within HIF-1α expressing cells based on differential expression of genes other than HIF-1α (Figure 5), uncovering other genes that may play a key role in hypoxic cellular processes. Milo allows researchers to validate these additional RNA targets revealed by single-cell RNA-seq experiments with protein expression data which may provide critical insights into the role these genes play in cellular function.

Download our application note to learn how Milo was used in parallel with single-cell RNA-seq at the Stanford Functional Genomics Facility to validate single-cell RNA expression studies with single-cell protein expression data.

Want to learn more about Milo?

Learn more about Milo by visiting the links below.