Pet Scans to Detect HIV Reservoirs - new research
Download PDF here
A live look at the AIDS virus
By Jon Cohen
9 March 2015
Pet scans to detect HIV reservoirs: ".....immunoPET might help clarify the power of different cure strategies. Many investigators are promoting the "shock and kill" idea, which attempts to prod reservoirs into making new copies of HIV, setting up the infected cells for destruction. "We could do shock and image."
"A positron emission tomography (PET) scan detected the antibodies and thus revealed exactly where the virus was hanging out inside the monkeys.....A positron emission tomography (PET) scan detected the antibodies and thus revealed exactly where the virus was hanging out inside the monkeys.....In another experiment that has practical implications for humans, Villinger and his colleagues gave three of the infected monkeys ARV drugs and then did the same kind of PET scans 5 weeks later. Although none of the monkeys had detectable SIV in their blood by the time of the scans, they all had SIV reproducing in multiple tissues......scans might help resolve a related long-standing debate in the field: Just how frequently do fully suppressed people continue producing low levels of the virus? "
"David Margolis, who has pioneered human studies that attempt to shock and kill reservoirs, says immunoPET is a "nice new tool" but cautions that it's unclear to him how well the signals it detects match viral levels in the body."
A live look at the AIDS virus
Nature March 2015 by Jon Cohen (full research paper follows below)
Show me the monkey.
Seeing is believing, and a study in rhesus macaques with a new imaging technique reveals for the first time a real-time map of an AIDS virus replicating in the entire body of a living animal. The results point to some unexpected hideouts of the simian AIDS virus, or SIV. And the experiments also show that when the monkeys are given antiretroviral (ARV) drugs, the amount of virus that persists differs by location in the body. The innovative tool promises to clarify the still-murky details of the initial infection process and may help guide drug, vaccine, and cure research in people.
"It's fantastic," says Thomas Hope, an immunologist at Northwestern University Feinberg School of Medicine in Chicago, Illinois, who investigates how HIV, the human counterpart to SIV, infects cells. "The whole monkey shows you things you can't comprehend by just looking at cells or biopsies of tissues."
To obtain this unprecedented look at an AIDS virus in action, a research team led by Francois Villinger of Emory University in Atlanta borrowed a technique used with cancer patients. The researchers began by attaching a radioactive molecule to an antibody that targets the surface protein of SIV. They then gave 12 monkeys chronically infected with SIV the harmless, engineered antibody, which sought out and latched onto the SIV in their bodies. A positron emission tomography (PET) scan detected the antibodies and thus revealed exactly where the virus was hanging out inside the monkeys. As expected from previous biopsy studies, the A positron emission tomography (PET) scan detected the antibodies and thus revealed exactly where the virus was hanging out inside the monkeys. But there were several startling findings.
The immunoPET method illuminated surprisingly high levels of the SIV antibody in the nasal cavity, the investigators report online today in Nature Methods. "The entire upper respiratory tract is rich with lymphatic tissue, and we just never thought of that," says Timothy Schacker of the University of Minnesota, Twin Cities, who studies how HIV causes disease. "That's a really cool finding."
Villinger was particularly struck by the high levels of SIV in the genital tract of males, given that sexual transmission of the AIDS virus does not occur all that readily. "The epididymis [the tubes that carry sperm] of the monkeys are just lighting up," Villinger says. "It's mind-boggling." The virus also had an unexpected penchant for the lung, an organ that has received relatively little attention from HIV researchers.
Hope says this tool could be used to clarify what takes place in monkeys, and presumably people, during the first weeks of an infection with an AIDS virus. "That sets the tone for what happens years later," he explains. "We need to know where the virus goes, how it gets there, and why it's a benefit to the virus. These pieces of the puzzle are really important."
In another experiment that has practical implications for humans, Villinger and his colleagues gave three of the infected monkeys ARV drugs and then did the same kind of PET scans 5 weeks later. Although none of the monkeys had detectable SIV in their blood by the time of the scans, they all had SIV reproducing in multiple tissues. This meshes with work by Schacker's group that found evidence of viral replication in biopsied tissue from "fully suppressed" HIV-infected people on treatment who have undetectable levels of the virus in their blood. Schacker's team has also shown that ARV drugs had difficulty penetrating these tissues. ImmunoPET scans could be an "excellent, more noninvasive way" of evaluating ways to get drugs to those hard-to-reach reservoirs, Schacker says.
If immunoPET can be adapted to tracking HIV in humans, which Villinger and others say is likely, scans might help resolve a related long-standing debate in the field: Just how frequently do fully suppressed people continue producing low levels of the virus? The evidence that antivirals have difficulty reaching some tissues convinces Schacker that it is a common phenomenon. "The drugs are not doing everything we think they're doing," he contends.
The other camp of researchers insists that treatment fully suppresses the virus and that it persists only because of "reservoirs" of long-lived cells that have HIV DNA woven into their chromosomes. Neither drugs nor the immune system reach this latent DNA, which sits poised to restart the viral fire at any time.
Figuring out a way to identify which tissues harbor these reservoirs in fully suppressed people tops the agenda of many HIV researchers, as it theoretically could fine-tune attempts to purge them and cure people. Villinger suggests immunoPET might help with this, too. "If you stop antiretroviral treatment, then you can see where the virus rekindles," he says.
In a similar vein, immunoPET might help clarify the power of different cure strategies. Many investigators are promoting the "shock and kill" idea, which attempts to prod reservoirs into making new copies of HIV, setting up the infected cells for destruction. "We could do shock and image," Villinger says, theoretically revealing the interventions that have the most impact.
David Margolis, who has pioneered human studies that attempt to shock and kill reservoirs, says immunoPET is a "nice new tool" but cautions that it's unclear to him how well the signals it detects match viral levels in the body. He also notes that a great deal of latent HIV DNA codes for defective, "dead-end" viruses that cannot cause infection themselves. If the labeled antibodies detect their presence, it could be misleading when it comes to evaluating whether reservoirs are shrinking.
Hope agrees that immunoPET tracking of AIDS viruses needs to be refined and still has technical hurdles to clear, but says the work advances the field. He even expressed surprise that it was in Nature Methods rather than the more prestigious mother journal, Nature. (Villinger says Nature reviewed and rejected the paper after a great deal of back-and-forth.) "It deserves to be there," Hope says. "This is groundbreaking."
Nature Methods March 2015
Whole-body immunoPET reveals active SIV dynamics in viremic and antiretroviral therapy-treated macaques
Philip J Santangelo1, Kenneth A Rogers2, Chiara Zurla1, Emmeline L Blanchard1, Sanjeev Gumber3,4, Karen Strait5, Fawn Connor-Stroud5, David M Schuster6, Praveen K Amancha2, Jung Joo Hong2, Siddappa N Byrareddy4, James A Hoxie7, Brani Vidakovic1, Aftab A Ansari4, Eric Hunter4 & Francois Villinger2,4
1Wallace H. Coulter Department of Biomedical Engineering, Georgia Institute of Technology and Emory University, Atlanta, Georgia, USA. 2Division of Microbiology and Immunology, Yerkes National Primate Research Center, Emory University, Atlanta, Georgia, USA. 3Division of Pathology, Yerkes National Primate Research Center, Emory University, Atlanta, Georgia, USA. 4Department of Pathology and Laboratory Medicine, Emory University School of Medicine, Atlanta, Georgia, USA. 5Division of Veterinary Medicine, Yerkes National Primate Research Center, Emory University, Atlanta, Georgia, USA. 6Department of Radiology and Imaging Sciences, Emory University School of Medicine, Atlanta, Georgia, USA. 7Department of Medicine, University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania, USA.
The detection of viral dynamics and localization in the context of controlled HIV infection remains a challenge and is limited to blood and biopsies. We developed a method to capture total-body simian immunodeficiency virus (SIV) replication using immunoPET (antibody-targeted positron emission tomography). The administration of a poly(ethylene glycol)-modified, 64Cu-labeled SIV Gp120-specific antibody led to readily detectable signals in the gastrointestinal and respiratory tract, lymphoid tissues and reproductive organs of viremic monkeys. Viral signals were reduced in aviremic antiretroviral-treated monkeys but detectable in colon, select lymph nodes, small bowel, nasal turbinates, the genital tract and lung. In elite controllers, virus was detected primarily in foci in the small bowel, select lymphoid areas and the male reproductive tract, as confirmed by quantitative reverse-transcription PCR (qRT-PCR) and immunohistochemistry. This real-time, in vivo viral imaging method has broad applications to the study of immunodeficiency virus pathogenesis, drug and vaccine development, and the potential for clinical translation.
Delineating viral replication in the context of generalized infections has traditionally relied on indirect measures, such as evaluating viral loads in plasma or via specific tissue biopsies. Such approaches have been valuable for the clinical management of viral infections, although they generally do not identify the site or source of virus replication in vivo. In a small percentage of HIV-infected individuals termed elite controllers (ECs), virus replication is controlled to undetectable levels without antiretroviral intervention, and disease progression may be delayed for decades1, 2. Despite undetectable virus in the plasma, virus evolution continues to occur consistently with ongoing tissue-contained virus replication3, 4. It is critical to identify tissue sites that can possibly serve as viral reservoirs so that the mechanisms by which such reservoirs are maintained can be identified. This would facilitate the development of strategies for eliminating these reservoirs, particularly in individuals treated with highly active antiretroviral therapy (ART)5. Ideally, a method to identify changes in virus localization would be minimally invasive as well as specific, sensitive and amenable to repeated application. Here we describe the application of whole-body imaging to the detection and localization of sites of SIV infection in chronically infected, ART-treated and EC macaques.
We describe the development of a non-invasive, sensitive immunoPET radiotracer and an approach to define the localization of SIV-infected tissue and free virus within live, chronically viremic, ART-treated and EC animals. The method can be repeated within the same animals (for example, before and during ART) without any adverse effect. In viremic animals, infection was concentrated within the mucosa of the gut, reiterating that these tissues are a major site of SIV replication12, 20, 21. However, we also observed discrete areas of virus replication, confirmed by qRT-PCR and IHC, both in nasal-associated tissues (post-ART) and in the reproductive tract of male animals. Within chronically infected, aviremic, ART-treated as well as EC animals, the methodology was able to detect residual virus, corroborated by qRT-PCR data. Thus, this approach provides the ability to identify novel areas of virus replication that may otherwise be difficult to sample in live animals. It may also provide a powerful tool to monitor the kinetics of viral replication in tissues over time during the application of novel therapeutic approaches. With the current efforts toward HIV eradication or functional cure, we believe that this method can be useful for determining organ-specific efficacy, which is crucial to the elimination of virally infected cells.
Our data also indicate that care must be taken when analyzing biopsies from aviremic subjects, especially ART-treated subjects, which may result in erroneous conclusions due to sampling (Supplementary Fig. 6 and Supplementary Table 3b). The detailed study of the cellular composition of these specific foci of infection, combined with site-specific drug metabolite levels and aided by the ability to image these specific sites, will likely be key to the development of directed therapies aimed at clearing infection from these sites in both controllers and individuals under ART28, 29, 30.
Although additional refinements to improve contrast and uptake are ongoing, we think that the methodology should be translatable to humans in the future because of the availability of anti-Env HIV antibodies31 and because the imaging approach is based on technologies already used in the clinic17, 18. It is applicable for studies investigating the eradication of HIV infection and targeting of virus reservoirs28, 32. Moreover, use of this technology during acute SIV infection may provide improved delineation of spatial kinetics of viral spread based on the route of infection and allow the identification of stages at which interruption of infection may be targeted using prophylactic methods33.
We hypothesized that the higher sensitivity and image quality of 64Cu-based positron emission tomography (PET) over that of single-photon emission computer tomography6 could be exploited to detect the major target tissues infected by SIV or HIV by scanning the entire body7, 8, 9, 10. To test this hypothesis, we employed the SIV-infected rhesus macaque as a model of pathogenic HIV infection11, 12 and chose the glycoprotein Gp120 as the in vivo target7.
Development of the immunoPET probe
We selected the SIV Env protein-specific monoclonal antibody (mAb) clone 7D3 as the basis of the probe because of its broad SIV Env specificity13, 14, 15. 7D3 binds the CCR5 binding site of Gp120 and prevents syncytia formation in vitro with SIVMACCP-MAC, although it does not affect soluble CD4 binding or neutralize SIVmac239 (ref. 13). Furthermore, three 7D3 molecules can bind to the trimeric Env of SIVMACCP-MAC and SIVMAC239 (ref. 15). To mitigate the immunogenicity of murine antibodies (probe antibodies were not detected after two administrations; data not shown), decrease nonspecific interactions and enable the chelation of 64Cu, the mAb was modified with linear 10-kDa poly(ethylene glycol) (PEG)16 through standard succinimidyl ester-amino chemistry and the chelator DOTA NHS (1,4,7,10-tetraazacyclododecane-1,4,7,10-tetraacetic acid mono), which has been demonstrated to be stable in humans and mice for at least 72 h (refs. 17,18).
We performed in vitro assays to validate the probe (Supplementary Fig. 1).
First, we incubated the 64Cu-labeled radiotracer with varying percentages of SIV1C cells and uninfected Hut78 cells (Supplementary Fig. 1a), demonstrating a linear binding response with decreasing percentages of infected cells, as measured with a gamma counter. Next we measured gamma counts from equal aliquots of cryopreserved and thawed splenocytes and lymph node cells, collected at necropsy from uninfected and SIV-infected animals (Supplementary Fig. 1b). The data demonstrate that even after a freeze-thaw cycle, the 7D3-PEG-64Cu-DOTA probe bound specifically to infected cells. Subsequently, we performed a competition assay whereby 'cold' 7D3-PEG-DOTA was used to compete with bound 'hot' 7D3-PEG-64Cu-DOTA from SIV1C and Hut78 cells, demonstrating epitope-specific binding of the probe (Supplementary Fig. 1c). In order to assess differences between 7D3-PEG-DOTA- and 7D3-PEG-DyLight 650-labeled probes (used in flow cytometry assessments), we used both probes to label SIV1C cells, demonstrating specific binding (Supplementary Fig. 1e). This experiment showed that the fluorescent probe binds with the same efficacy as the 7D3-PEG-DOTA (cold) probe. We then assessed the DyLight 650-labeled probe, using flow cytometry for its binding to SIV1C cells in the presence of serum and SIV-specific antibodies generated during infection (Supplementary Fig. 1f).
Although there was less binding of 7D3 in the presence of the pre- and post-infection serum than with monkey serum absent, we noted specific binding to SIV1C relative to uninfected Hut78 cells.
Characterization of chronic SIV infection
To generate images of SIV dissemination in vivo, we injected the 7D3-PEG-64Cu-DOTA probe (labeled with 1-3.5 mCi/mg 64Cu) into rhesus macaques intravenously. We tested probe stability by pelleting virus from the plasma of the infected and uninfected animals and measuring the radioactivity (Supplementary Fig. 1d). These data demonstrate that specific probe binding to virus was still possible 24 h after injection of the probe. We found PET and computed tomography (CT) imaging (hereafter PET/CT)in both viremic and control animals to be optimal 24-36 h after injection (data not shown). We then imaged macaques chronically infected with SIVMAC239 and uninfected control animals after injection with either a 64Cu-labeled, PEG-modified 7D3 mAb or a similarly labeled control IgG to test the ability of the modified mAb to specifically target and detect SIV-infected cells and tissues in vivo, displaying frontal, sagittal and axial images (Fig. 1a,b, Supplementary Fig. 2a-f and Supplementary Table 1). We generated standardized uptake value (SUV) maps for each PET and corresponding CT image using OsiriX software (Fig. 1a,b and Supplementary Fig. 2a-f). SUV is essentially the measured activity within a region of interest, normalized by the radioactive dose divided by the subject's weight. In the viremic animals, SUV values greater than those observed in the control cases were localized to the gastrointestinal (GI) system, specifically within the ileum, jejunum and colon, and the axillary and inguinal lymph nodes, correlating with previously reported findings12, 19, 20, 21. We also detected uptake within the lungs, a less well-characterized site of viral infection22. As can be seen in detail in slice sequences for monkeys Viremic 1 and 2, the uptake followed the contours of the ileum and colon (Supplementary Fig. 3a,b). We also consistently detected antibody uptake within the nasal cavity, likely reflective of the nasal-associated lymphoid tissue (NALT) or nasal turbinates (Fig. 1a), an area that has received little attention as a source of viral replication. In males, we frequently observed uptake within the genital tract, specifically in the vas deferens and epididymis, corroborating previous reports23, 24, 25. In contrast to a previous report26, infection of epithelial cells was not detected in the male genital organs. In the uninfected monkeys, background was evident within the liver, heart, kidneys and spleen, which is typical of antibody-based radiotracers17, 18. It should be noted, however, that SUVs were significantly higher in infected animals than in control cases in each of these organ systems, indicating specific uptake (P < 0.05, Kruskal-Wallis test; Supplementary Table 2). A sequence of frontal slice images obtained on the uninfected monkey Control 1 after administration of 64Cu-7D3 demonstrates that uptake was minimal throughout the GI tract and that background was restricted to the heart, liver, kidneys and spleen (Supplementary Fig. 3c)
PET quantification and comparison with qRT-PCR and IHC
We verified the imaging signal data by qRT-PCR and through the examination of sections of the specific tissue of interest using immunohistochemistry (IHC) for SIV Gag protein (Fig. 1c,d and Supplementary Fig. 2). These confirmatory studies used rectal biopsy tissues obtained immediately after imaging and/or tissues obtained post-mortem. On the basis of the IHC data, the GI tract, lymph nodes and spleen all contained infiltrating SIV-infected cells of lymphocyte or macrophage morphology. Negative control tissue (Fig. 1d) did not contain any detectable signal using the same IHC protocol. We also performed qRT-PCR using tissue samples from the colon, small bowel, spleen and inguinal and axillary lymph nodes for all of the chronically infected animals and controls (Fig. 1e). Corroborating our IHC results, we detected viral RNA in all cases, with the highest levels in the colon; we did not detect RNA in the controls. In addition, we quantified the 64Cu-radioactivity associated with aliquots of rectal biopsies from viremic monkeys Viremic 1 and 2 and uninfected Controls 1 and 2 with a gamma counter. The signal from these tissues, normalized for the mass of the biopsy and the total amount of radioactivity administered, was 17.6 times higher on average in infected animals (Fig. 1f) than in both controls, providing additional confirmation of the specificity of the PET imaging.
To compare the PET results from viremic animals to those of the controls, we quantified the data using SUV (Fig. 1a,b,e,g and Supplementary Fig. 2). We chose volumes of interest in the PET/CT fusion images by outlining the organ manually in the CT images throughout the image slices. Using the organ volume as the region of interest (ROI), we then determined the maximum SUV within that organ and compared the SUVmax within viremic and uninfected animals with the qRT-PCR results from the corresponding colon, small bowel, spleen and inguinal and axillary lymph nodes of the same animal (Fig. 1e). These data suggest that the PET SUVmax values mimic the general trends of the PCR data. For the spleen, which tends to have higher background uptake, the SUVmax minus background is a more relevant comparison with the qRT-PCR data. It should be pointed out that the SUV data are unlikely to match the PCR data precisely, as protein and RNA expression levels may differ.
Next, to address the specificity of the PET signals, we compared the SUVmax measurements for various tissues from chronically infected and uninfected macaques injected with either the modified 7D3-labeled antibody or with a labeled isotype control antibody (Fig. 1g and Supplementary Fig. 2g). The data suggest that there is increased uptake within organ systems likely to contain virus or virally infected cells and tissue. For each group of animals, we then performed an overall comparison of the PET measurements that included the signals of all organs, applying a fully nested hierarchical ANOVA model for the SUVmax response. The null hypothesis for the comparison was that the viral status of the animal and the injected probe would not influence the imaging results. We found that the viremic animals were significantly different from the uninfected controls (P = 7.39 x 10-6). However, animals within each group, infected or uninfected, were not significantly different from each other (P = 0.89). The analysis also showed that both the infection status (animal group) and organs contributed significantly to the SUVmax value. We found that for the chronically viremic animals and both control groups, the measurements for each organ within each group were significantly different from each other, yielding P values of 1.4 x 10-9, 7.1 x 10-27 and 4.78 x 10-19, respectively. Furthermore, when we applied the Kruskal-Wallis test for each organ separately (Supplementary Table 2), the signals measured in viremic and uninfected controls were statistically distinct except for in muscle.
Another method of assessing specific uptake for a particular organ is to examine the dynamics of uptake. Owing to logistics, we scanned animals at 12, 24 and/or 36 h after injection and plotted the ratios of the average SUVmax values at each time point for each organ (Fig. 1h). When we compared the viremic monkeys with the aviremic controls, all of the ratios in viremic animals were higher-typically above 0.6, with the GI tract giving values >1.0-indicating continued specific uptake of the probe. The decay rate and SUVmax values in whole blood (SUVmax ≈ 0.3-0.4; data not shown) were similar to those of muscle, and remained similar between groups. We did not detect uptake in the central nervous system, probably because of probe exclusion by the blood-brain barrier27.
SIV localization before and during antiretroviral therapy
In order to confirm the sensitivity of this method and its ability to track SIV replication anatomically during treatment, we first imaged three chronically infected animals (ART 1, ART 2 and ART 3) 36 h after injection with our modified 7D3 probe and then initiated them on ART (20 mg PMPA (9-(2-phosphonomethoxypropyl)adenine) per kg body weight per day (mg per kg per d) and 50 mg per kg per d emtricitabine (FTC) each subcutaneously and 100 mg/d for 40 d of integrase inhibitor L-870812) (Fig. 2 and Supplementary Fig. 4). All three animals were aviremic by 3-4 weeks of treatment (having viral loads <60 copies of RNA per ml; Supplementary Fig. 5) and imaged again after 5 weeks on ART (Fig. 2 and Supplementary Fig. 4). Prior to treatment, there was measurable SIV signal localized within the GI tract, NALT, genital tract and axillary and inguinal lymphoid tissue. After 34 d of treatment, all organ systems exhibited decreased uptake (Fig. 2a,b, Supplementary Fig. 4 and Supplementary Table 3a).
However, there was residual signal (above the background) in all organ systems, with moderate SUVmax values still remaining in the colon, spleen, male genital tract, NALT and individual lymph nodes for specific animals. In none of the cases did the SUVmax decrease to our measurable limit (background). To assess the statistical significance of the SUVmax measurements, we performed a hierarchical ANOVA analysis as described above. The differences between the SUVmax data before and after treatment from all of the organs imaged were significant, with a P value of 0. In a pairwise comparison, ART 1 and ART 3 were significantly different from ART 2 (P = 0.0027), demonstrating the individual variation in ART treatment.
To verify the imaging results, we performed qRT-PCR on multiple tissue samples and compared the data directly with SUVmax data (Fig. 2c and Supplementary Fig. 4c). The samples included colon, small bowel, spleen, and right and left inguinal and axillary lymph nodes; these were collected at necropsy performed after 39 or 40 d on ART. Even though our PET procedure measures Env protein and qRT-PCR detects viral RNA, residual virus or infected cells were indeed present in the locations identified by PET. Both the spatial variation within an animal and the variation between animals suggested by PET was confirmed with qRT-PCR data (Supplementary Fig. 6 and Supplementary Table 3b), with two-orders-of-magnitude variation within an organ and between animals. Additionally, the nasal turbinate, genital tract and lung samples were all positive for viral RNA (Supplementary Table 3b), indicating virus localization during both chronic and treated conditions, a similar result to that observed in viremic animals.
SIV localization in elite controllers
We then applied the methodology to SIV-infected ECs. ECs are individuals that naturally suppress SIV (or HIV) replication to undetectable levels in plasma for extended periods of time without antiretroviral intervention1, 2. Given the challenges studying viral persistence in these animals, they were an ideal test for our approach. EC monkeys exhibited detectable uptake (Fig. 3 and Supplementary Fig. 7) within the GI tract, genital tract, NALT, lungs, spleen and axillary lymph nodes. These imaging data were supported by IHC in biopsy samples (Supplementary Fig. 7, Supplementary Table 2 and Supplementary Note). In ECs, the uptake was restricted to smaller regions or foci as compared to viremic animals. When quantified, the SUVmax organ signal appeared to approximate the results in viremic animals (Fig. 3b). However, when we applied a hierarchical ANOVA, we found that overall the PET SUVmax data for the viremic animals were statistically distinct from those of the ECs and the control animals (P = 2.78 x 10-5). To clarify the differences between the ECs and viremic animals, we measured the SUVmean within the GI tract and compared the voxel fractions (fraction of total volume of GI tract) (Fig. 4 and Supplementary Fig. 8). The GI tract values in the viremic animals were 2.1 and 6.38 times greater than in the EC animals for SUVmean and voxel fraction, respectively. Thus, although the viremic macaques and ECs had regions of comparably high uptake, in ECs, this was spatially restricted to much smaller volumes, and therefore the overall probe uptake was lower. We calculated additional metrics quantifying the spatial distributions within the GI tract, which further supported this conclusion (Fig. 4a,b and Supplementary Note).