Abstract Body:

Introduction 

Despite NMR spectroscopy being a powerful tool for structural and chemical characterization, it suffers from low sensitivity, hence conventionally needing large sample volumes at relatively high concentration. To limit these requirements, miniaturization of NMR probes has been explored, exploiting the inverse proportionality of the SNR with the diameter of the coil within the radiofrequency skin depth regime.[1, 2] However, challenges arise when coils of sensitive volumes of 1μL or below are integrated in the NMR circuit, while aiming at detecting limited sample amounts and their dynamic change. A field that requires optimizing this regime is cancer biology, where the formation/depletion of certain metabolites by a limited number of live cells is sought as a probe to assess cancer metabolism and response to treatments.[3, 4] Previous efforts developed a micro-scale NMR platform of a μL-size sensitive region, detecting metabolic flux from approx. 10⁵ live cells.[5-7] In this work we present a double ¹H/¹³C resonance NMR probe consisting of a solenoid coil with 100nL sensitive region. A suspension of live cells and hyperpolarized (HP) [1-¹³C]pyruvate are loaded in the coil, and dynamic changes in pyruvate/lactate concentrations by fractions of femtomoles are detected from just 2000 live cells at a time. Through an integrated microfluidic channel, the probe is used as high-throughput platform to perform non-destructive quantitative analysis of metabolic flux of different leukemia cell lines.

Methods

A solenoid copper coil of 100nL inner volume was fabricated and embedded in a PDMS mold to create a microfluidic channel across the solenoid (Figure 1A). The coil was integrated in a circuit and connected to two channels capacitance-adjusted to be resonant with the ¹H and ¹³C Larmor frequencies at 3T. The analytical protocol is as follows: i) A cell suspension was prepared immediately before the start of the NMR measurement. ii) [1-¹³C]pyruvate hyperpolarized using DNP was mixed with the suspension, and the first aliquot, containing 2000 cells, was loaded into the coil through the microfluidic channel. iii) The probe was positioned inside the 3T MRI bore, and the NMR acquisition started. iv) The following NMR measurement consisted of a series of 45-degree single pulses, each followed by FID detection, until the ¹³C NMR signal had fully decayed. v) A successive aliquot of the original suspension was loaded in the coil, and the same NMR scheme was performed. Step v) was repeated iteratively over the following 60-to-120 seconds. The NMR peaks of [1-¹³C]pyruvate and [1-¹³C]lactate were integrated over time and their ratio, normalized by the number of cells and the moles of starting pyruvate, gave us a measure of the metabolic flux of the live cells. 

Results and Discussion 

First we determined if the system could robustly measure metabolic flux in cells. Figure 1B demonstrates time dependent metabolic conversion of 9 samples, each containing 2000 cells. To test the ability of the nanocoil to measure a biologically relevant metabolic flux difference from 2000 live cells, two leukemia lines were studied, MOLM13 and NOMO1, the latter known to produce approximately 2-fold more lactate from HP pyruvate. Additionally, MOLM13 cells were treated with ABT199, which alters NAD+ and NADH levels resulting in an increase of pyruvate-to-lactate flux. Our results for both MOLM13 vs NOMO1 (Figure 1C-E) and untreated MOLM13 vs MOLM13+ABT199 (Figure 1F) reproduce the flux values reported in the literature, this time measured with 1-to-3 orders of magnitude fewer cells than previous studies. Taken together, this work demonstrates the development and application of an HP nano-NMR platform for the study of live cell metabolism in mass limited samples, more than an order of magnitude more sensitive than current approaches, considerably improving the otherwise demanding analysis of scarce samples, such as primary cancer cells.

Author