AI on a chip Isozaki, Akihiro; Harmon, Jeffrey; Zhou, Yuqi ...
Lab on a chip,
08/2020, Letnik:
2, Številka:
17
Journal Article
Recenzirano
Artificial intelligence (AI) has dramatically changed the landscape of science, industry, defence, and medicine in the last several years. Supported by considerably enhanced computational power and ...cloud storage, the field of AI has shifted from mostly theoretical studies in the discipline of computer science to diverse real-life applications such as drug design, material discovery, speech recognition, self-driving cars, advertising, finance, medical imaging, and astronomical observation, where AI-produced outcomes have been proven to be comparable or even superior to the performance of human experts. In these applications, what is essentially important for the development of AI is the data needed for machine learning. Despite its prominent importance, the very first process of the AI development, namely data collection and data preparation, is typically the most laborious task and is often a limiting factor of constructing functional AI algorithms. Lab-on-a-chip technology, in particular microfluidics, is a powerful platform for both the construction and implementation of AI in a large-scale, cost-effective, high-throughput, automated, and multiplexed manner, thereby overcoming the above bottleneck. On this platform, high-throughput imaging is a critical tool as it can generate high-content information (
e.g.
, size, shape, structure, composition, interaction) of objects on a large scale. High-throughput imaging can also be paired with sorting and DNA/RNA sequencing to conduct a massive survey of phenotype-genotype relations whose data is too complex to analyze with traditional computational tools, but is analyzable with the power of AI. In addition to its function as a data provider, lab-on-a-chip technology can also be employed to implement the developed AI for accurate identification, characterization, classification, and prediction of objects in mixed, heterogeneous, or unknown samples. In this review article, motivated by the excellent synergy between AI and lab-on-a-chip technology, we outline fundamental elements, recent advances, future challenges, and emerging opportunities of AI with lab-on-a-chip technology or "AI on a chip" for short.
The excellent synergy between artificial intelligence and lab-on-a-chip technology is described with applications.
Mesenchymal stem cells (MSCs) have a tumor-homing ability—they accumulate inside tumors after systemic injection, and may thus be useful as carriers for tumor-targeting therapy. To use MSCs ...effectively as an anti-cancer therapy, they must first be functionalized with a large amount of anti-cancer drugs without causing any significant changes to their tumor-tropism. In the present study, we attempted to modify the cell surface of MSCs with doxorubicin-loaded liposomes (DOX-Lips), using the avidin-biotin complex method, and evaluated delivery efficiency and anti-tumor efficacy of DOX-Lip-modified MSCs. The amount of DOX in DOX-Lip-modified C3H10T1/2 cells, a murine mesenchymal stem cell line, was approximately 21.5 pg per cell, with no significant changes to the tumor-tropism of C3H10T1/2 cells. Notably, DOX-Lip-modified C3H10T1/2 cells significantly suppressed the proliferation of firefly luciferase-expressing murine colon adenocarcinoma colon26/fluc cells, compared to DOX-Lips alone. Fluorescent DOX accumulated at the cell contact surface and inside green fluorescence protein-expressing colon26 (colon26/GFP) in co-cultures of DOX-Lip-modified C3H10T1/2 and colon26/GFP cells. This localized distribution was not observed when only DOX-Lips was added to colon26/GFP cells. These results suggest that DOX-Lips are efficiently delivered from DOX-Lip-modified C3H10T1/2 cells to the neighboring colon26 cells. Furthermore, DOX-Lip-modified C3H10T1/2 cells suppressed tumor growth in subcutaneous tumor-bearing mice, and in a lung metastasis mouse model. Taken together, these results indicate that the intercellular delivery of DOX may be enhanced using DOX-Lip-modified MSCs as an efficient carrier system for targeted tumor therapy.
Display omitted
Management of phenylketonuria (PKU) requires lifelong restriction of phenylalanine (Phe) intake using specialized medical foods to prevent neurocognitive impairment in affected patients. However, ...dietary adherence is challenging to maintain while ensuring adequate nutrition, which can lead to sub-optimal clinical outcomes. Metabolomics offers a systematic approach to identify new biomarkers of disease progression in PKU when using urine as a surrogate for blood specimens that is more accurate than self-reported diet records. Herein, the plasma and urine metabolome of a cohort of classic PKU patients (median age = 11 years;
n
= 22) mainly prescribed (78%) a Phe-restricted diet were characterized using multisegment injection-capillary electrophoresis-mass spectrometry (MSI-CE-MS). Overall, there was good mutual agreement between plasma Phe and tyrosine (Tyr) concentrations measured from PKU patients when using an amino acid analyzer based on UPLC-UV as compared to MSI-CE-MS with a mean bias of 12% (
n
= 82). Longitudinal measurements of recently diagnosed PKU infants (
n
= 3) revealed good long-term regulation of blood Phe with dietary management, and only occasional episodes exceeding the recommended therapeutic range (>360 μM) unlike older PKU patients. Plasma metabolomic studies demonstrated that non-adherent PKU patients had lower circulating concentrations of Tyr, arginine, 2-aminobutyric acid, and propionylcarnitine (
q
< 0.05, FDR) that were inversely correlated to Phe (
r
−0.600 to −0.830). Nontargeted metabolite profiling also revealed urinary biomarkers associated with poor dietary adherence among PKU patients, including elevated concentrations of catabolites indicative of Phe intoxication (
e.g.
, phenylpyruvic acid, phenylacetylglutamine, hydroxyphenylacetic acid). Additionally, PKU patients with poor blood Phe control had lower excretion of urinary compounds derived from co-metabolism of Tyr due to microbiota activity (
e.g.
, cresol sulfate, phenylsulfate), as well as several metabolites associated with inadequate nutrient intake, including low carnitine and B vitamin status (
e.g.
, folic acid, vitamin B12). Interestingly, an unknown urinary metabolite was strongly correlated with Phe excretion in PKU patients (
r
= 0.861), which was subsequently identified as imidazole lactic acid when using high resolution MS/MS. Overall, urine profiling offers a non-invasive approach for better treatment monitoring of individual PKU patients, which can also guide the design of novel therapies that improve adherence to Phe-restricted diets without acquired nutritional deficiencies.
Improved treatment monitoring of PKU patients: urinary biomarkers of dietary adherence and nutritional status.
Salmonella enterica serovar Dublin is a cattle-adapted serovar, and some infected cattle can become asymptomatic carriers. Identification of carrier animals is important for preventing the spread of ...infection within a farm, but low diagnostic sensitivity of the fecal culture method is problematic. In this study, we investigated isolation methods of four S. enterica Dublin strains. Selective enrichment using the tetrathionate broth showed better performance than Rappaport-Vassiliadis R10 broth, but one of the strains was not detectable. Since isolation of such strains by selective enrichment can be difficult, we designed a method using immuno-plates that concentrates S. enterica Dublin by antigen-antibody reaction. Our method is able to detect approximately 200 clony-forming units of S. enterica Dublin in 0.1 g of cattle feces. If S. enterica Dublin was isolated from cattle with clinical signs, the method to identify carriers in the farm should be based on the growth kinetics of the target S. enterica Dublin strain.
Abstract
A few final documents adopted by the Review Conferences of the Nuclear Non-Proliferation Treaty ('NPT') include important references to nuclear disarmament and to Article VI of the NPT. It ...is frequently assumed or claimed that these references amount to subsequent agreements as to the interpretation of Article VI. However, such a view is difficult to maintain. With the help of the 2018 Draft Conclusions on subsequent agreements and practice, this article seeks to establish two conditions that such final documents ought to satisfy if they are to be treated as subsequent agreements under Article 31 of the Vienna Convention on the Law of Treaties ('VCLT'). The analysis suggests that the final documents in question do not satisfy such conditions. Based on this evaluation, the article also explores other theoretical frameworks both within and without treaty interpretation rules, that can adequately take into account the final documents of the NPT Review Conferences in discussing nuclear disarmament in international law.
Celotno besedilo
Dostopno za:
DOBA, IZUM, KILJ, NUK, ODKLJ, PILJ, PNG, PRFLJ, SAZU, UILJ, UKNU, UL, UM, UPUK
Intelligent image‐activated cell sorting (iIACS) has enabled high‐throughput image‐based sorting of single cells with artificial intelligence (AI) algorithms. This AI‐on‐a‐chip technology combines ...fluorescence microscopy, AI‐based image processing, sort‐timing prediction, and cell sorting. Sort‐timing prediction is particularly essential due to the latency on the order of milliseconds between image acquisition and sort actuation, during which image processing is performed. The long latency amplifies the effects of the fluctuations in the flow speed of cells, leading to fluctuation and uncertainty in the arrival time of cells at the sort point on the microfluidic chip. To compensate for this fluctuation, iIACS measures the flow speed of each cell upstream, predicts the arrival timing of the cell at the sort point, and activates the actuation of the cell sorter appropriately. Here, we propose and demonstrate a machine learning technique to increase the accuracy of the sort‐timing prediction that would allow for the improvement of sort event rate, yield, and purity. Specifically, we trained an algorithm to predict the sort timing for morphologically heterogeneous budding yeast cells. The algorithm we developed used cell morphology, position, and flow speed as inputs for prediction and achieved 41.5% lower prediction error compared to the previously employed method based solely on flow speed. As a result, our technique would allow for an increase in the sort event rate of iIACS by a factor of ~2.
Sort‐timing prediction is particularly essential for intelligent image‐activated cell sorting (iIACS) in order to achieve sorting at a high event rate. We propose and demonstrate a machine‐learning technique to increase the accuracy of sort‐timing prediction by taking into account cell morphology, position, and flow speed. We use timing data and images from morphologically heterogeneous budding yeast cells to assess our method and show the predicted improvement of event rate, yield, and purity.
•We modeled the spatial regulation of resource allocation.•We focused on root system architecture in response to nutrient availability.•We showed that local control is sufficient to adaptation to ...low-level nutrition.•In contrast, systemic control is required for adaptation to high-level nutrition.•Systemic control could be evolved when an excess amount of nutrition is present.
It is critical for a living organism to appropriately allocate resources among its organs, or within a specific organ, because available resources are generally limited. For example, in response to the nutritional environment of their soil, plants regulate resource allocation in their roots in order to plastically change their root system architecture (RSA) for efficiently absorbing nutrients. However, it is still not understood why and how RSA is adaptively controlled. Therefore, we modeled and investigated the spatial regulation of resource allocation, focusing on RSA in response to nutrient availability, and provided analytical solutions to the optimal strategy in the case of simple fitness functions. We first showed that our model could explain the experimental evidence where root growth is maximized at the optimal nutrient concentration under the homogeneous condition. Next, we extended our model to incorporate the spatial heterogeneity of nutrient availability. This extended model revealed that growth suppression by systemic control is required for adapting to high nutrient conditions, whereas growth promotion by local control is sufficient for adaptation to low-nutrient environments. This evidence predicts that systemic control can be evolved in the presence of excessive amounts of nutrition, consistent with the ‘N-supply’ systemic signal that is observed experimentally. Furthermore, our model can also explain various experimental results using nitrogen nutrition. Our model provides a theoretical basis for understanding the spatial regulation of adaptive resource allocation in response to nutritional environment.
In northeastern Japan, an area of high precipitation and mountains, beech (
Fagus creanata
Blume), larch (
Larix kaempferi
Lamb.), cedar (
Cryptomeria japonica
D. Don) and black locust (
Robinia ...pseudoacacia
L.) were evaluated for N resorption and N isotope fractionation in pre- and post-abscission leaves in comparison to green leaves. The highest leaf N concentration in summer corresponded to the N-fixing black locust, followed in decreasing order by the deciduous beech and larch and evergreen cedar. On the other hand, the lowest N resorption efficiency corresponded to black locust and the highest to beech, in increasing order by larch and cedar. All tree species returned significant amounts of N before leaf abscission; however, N isotope fractionation during leaf N resorption was only found for beech, with a depleted N isotope value from green to pre-abscission leaf. The most N, however, was resorbed from pre-abscission to post-abscission. This result may indicate that δ
15
N fractionation took place during N transformation processes, such as protein hydrolysis, when the concentration of free amino acids increased sharply. The difference in the type of amino acid produced by each species could have influenced the N isotope ratio in beech but not in the other tree species. The results of this study showed that it is possible to infer the type and timing of processes relevant to N resorption by analyzing leaf δ
15
N variation during senescence.
Main conclusion
The local and long-distance signaling pathways mediated by the leucine-rich repeat receptor kinase HAR1 suppress root branching and promote primary root length in response to nitrate ...supply.
The root morphology of higher plants changes plastically to effectively absorb nutrients and water from the soil. In particular, legumes develop root organ nodules, in which symbiotic rhizobia fix atmospheric nitrogen in nitrogen-poor environments. The number of nodules formed in roots is negatively regulated by a long-distance signaling pathway that travels through shoots called autoregulation of nodulation (AON). In the model plant
Lotus japonicus
, defects in AON genes, such as a leucine-rich repeat receptor kinase
HYPERNODULATION ABERRANT ROOT FORMATION 1
(
HAR1
), an orthologue of
CLAVATA1
, and the F-box protein
TOO MUCH LOVE
(
TML
), induce the formation of an excess number of nodules. The loss-of-function mutant of HAR1 exhibits a short and bushy root phenotype in the absence of rhizobia. We show that the
har1
mutant exhibits high nitrate sensitivity during root development. The uninfected
har1
mutant significantly increased lateral root number and reduced primary root length in the presence of 3 mM nitrate, compared with the wild-type and
tml
mutant. Grafting experiments indicated that local and long-distance signaling pathways via root- and shoot-acting HAR1 additively regulated root morphology under the moderate nitrate concentrations. These findings allow us to propose that HAR1-mediated signaling pathways control the root system architecture by suppressing lateral root branching and promoting primary root elongation in response to nitrate availability.
The resistance of endometriotic tissue to progesterone can be explained by alterations in the distribution of progesterone receptor (PR) and estrogen receptor (ER) isoforms. The aims of this study ...were to examine the expressions of PR-A, PR-B, ERα and ERβ in endometrioma and assess whether these expressions are affected by dienogest or leuprolide acetate (LA) treatment.
We enrolled 60 females, including 43 patients with endometriosis (14 who received no medical treatment, 13 who received dienogest and 16 who received LA before undergoing laparoscopic surgery) and 17 patients with leiomyoma. The expression levels of PR and ER isoforms in eutopic and ectopic endometrium were assayed with quantitative real-time PCR, and confirmed with immunohistochemistry.
A decreased PR-B/PR-A ratio and an increased ERβ/ERα ratio were demonstrated in ectopic endometrium derived from females with endometriosis compared with the ratios observed in eutopic endometrium obtained from females without endometriosis. Although LA treatment did not affect the PR-B/PR-A and ERβ/ERα ratios, dienogest treatment increased the PR-B/PR-A ratio and decreased the ERβ/ERα ratio in patients with endometriomas.
Dienogest may improve progesterone resistance in endometriotic tissue by increasing the relative expressions of PR-B and PR-A, and decreasing the relative expressions of ERβ and ERα.