Zhao et al. J Transl Med (2020) 18:168 https://doi.org/10.1186/s12967-020-02335-7 Journal of Translational Medicine

REVIEW Open Access Microfuidic devices for neutrophil chemotaxis studies Wenjie Zhao1,2, Haiping Zhao3†, Mingxiao Li1* and Chengjun Huang1,2

Abstract Neutrophil chemotaxis plays a vital role in human immune system. Compared with traditional cell migration assays, the emergence of microfuidics provides a new research platform of cell chemotaxis study due to the advantages of visualization, precise control of chemical gradient, and small consumption of reagents. A series of microfuidic devices have been fabricated to study the behavior of neutrophils exposed on controlled, stable, and complex profles of chemical concentration gradients. In addition, microfuidic technology ofers a promising way to integrate the other functions, such as cell culture, separation and analysis into a single chip. Therefore, an overview of recent develop- ments in microfuidic-based neutrophil chemotaxis studies is presented. Meanwhile, the strength and drawbacks of these devices are compared. Keywords: Microfuidic, Neutrophil chemotaxis, Chemical gradient, Lab on a chip

Introduction velocity is about 1–20 µm/min, which varies with difer- Chemotaxis, the direct movement of cells along the ent chemokines and diferent gradients. Te directional chemical gradient, is crucial in biologic process, such as persistence means the ratio of relative displacement along innate immunity and cancer metastasis [1, 2]. Neutro- the gradient to total path length, which is infuenced by phils, which are the most numerous and most important the concentration of chemokines [4]. Sensory G-protein- cellular component of the innate immune response, are coupled receptors on neutrophil membranes can detect considered as the frst protective barrier against purulent slight concentration change of chemokines. In certain infections [3]. Neutrophils can persist for hours to days concentration ranges, the neutrophils even response to within the vasculature until they reach senescence. Tey a concentration diference between across their dimen- accumulate rapidly and efciently within minutes to areas sions of ~ 1% [5]. Mesenchymal stem cells (MSCs) have of infammation, which depends on the extreme sensitiv- the potential to diferentiate into a wide variety of other ity of the inciting stimuli. For this efective response, they cell types. Te migration of MSCs was induced by hom- can detect extracellular chemical gradients and move ing signals released by cells at the site of injury and/or towards higher concentrations. Te migration of these infammation [6]. Chemokines, cytokines, and growth cells is mediated by chemotaxis, which acts as the attrac- factors (such as IL-6 (interleukin-6) and PDGF (plate- tive force to determine the direction in which neutrophils let derived growth factor)) released from tissue damage move. Velocity and directional persistence are usually or apoptosis mobilize and recruit stem cells to the dam- used to characterize neutrophil chemotaxis. Te typical aged site, where they proliferate and diferentiate, eventu- ally replacing the damaged tissues [6]. Te factors induce upregulation of selectins and activation of integrins on *Correspondence: [email protected] the stem cell surface, enabling cells to interact with the † Haiping Zhao—Co-frst author endothelium. Stem cells subsequently adhere and trans- 1 Institute of Microelectronics, Chinese Academy of Sciences, 3 Beitucheng West Road, Beijing 100029, China migrate across the endothelial layer into tissues. In addi- Full list of author information is available at the end of the article tion, they are also sensitive to the physical properties of

© The Author(s) 2020. This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativeco​ ​ mmons.org/licen​ ses/by/4.0/​ . The Creative Commons Public Domain Dedication waiver (http://creativeco​ mmons​ .org/publi​ cdoma​ in/​ zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Zhao et al. J Transl Med (2020) 18:168 Page 2 of 19

the extracellular matrix (ECM) including stifness, topog- receptor-mediated uptake into degradatory vesicles raphy, and dimensionality [7]. Neutrophils are terminally (growth factor breakdown), dummy or atypical recep- diferentiated cells and cannot be genetically manipu- tors which are used to scavenge ligand or extracellular lated. A model system of neutrophil is the promyelocytic enzymes secreted or bound to the outer leafet of the leukemia cells line (HL-60) which can be induced to dif- cell membrane. In addition, chemoattractant breakdown ferentiate into neutrophil-like cells by using dimethyl sul- is important for sharpening gradients and bringing the foxide (DMSO) or dibutyryl-cAMP [8]. Te capacity of attractant levels down below the receptors’ dissociation diferentiated HL-60 cells to responds with chemokine- constant (Kd) so they can precisely perceived [15]. Even sis and chemotaxis to stimuli is diferent to neutrophils. in common assays, where cells are directly exposed to For example, polymorphonuclear cells (PMN)-like HL-60 gradients, chemoattractant breakdown makes chemot- cells respond diferently to thrombospondin (TSP) than axis markedly more efcient. human peripheral blood PMN. HL-60 cells were needed Since the concept of chemotaxis was proposed in to be diferentiated with dimethyl sulfoxide (DMSO), the 1960s, cell chemotaxis studies have been greatly retinoic acid (RA), vitamin D, or l-ascorbic acid (l-AA) improved. Traditional techniques in cell chemotaxis before chemokinesis and chemotaxis of HL-cells were researches based on cell culture in vitro are very mature assayed [9]. Undiferentiated HL-60 cells did not adhere now. One of the most widely used chemotaxis equipment and were not motile in response to TSP. With diferen- is “Boyden Chamber” [16] developed by Boyden in 1962, tiation, a maximal response was obtained with 100 to which is usually used to detect the chemotaxis of leuko- 300 nM TSP, tenfold lower than required for maximal cytes and macrophages. However, the Boyden Chamber PMN chemotaxis [10]. Tese diferences may refect (1) was unsuitable for observing single cell responses with an aberration in HL-60 diferentiation refecting their the unstable chemical gradient profles, which is also leukemic phenotype (2) diferentiation of HL-60 cells to hard to distinguish between chemotaxis and enhanced a cell type characteristic of “activated” PMN [10]. Te motility of cells. To observe cell migration in real time, biological mechanism of neutrophil chemotaxis is highly researchers developed “Zigmond Chamber” [5] based complex, including four steps [11]: neutrophil protec- on “Boyden Chamber”, which is considered as a proto- tion, adhesion to endothelial cells, endothelial cell wall type device based on microfuidic technology. Zigmond penetration, and migration to infected tissue. Incorrectly Chamber was the frst device to allow direct visualiza- signaled neutrophil chemotaxis can cause series cellular tion of cell behavior in the presence of the biomolecule malfunctions such as autoimmune diseases and fatal dis- gradient. Te main limitation of Zigmond Chamber was orders [12]. Terefore, the study of neutrophil chemot- that the characteristics of the gradient was determined by axis is of great signifcance in clinical medicine. the geometry of the device and the difusion coefcient Te classical chemotaxis model is defned as some of the biomolecule. Te generated gradient had short life source releases attractant (or repellent) into the envi- spans (~ 1 h) and was extremely sensitive to evaporation. ronment, while a distant sink absorbs it, forming a dif- Ten, “Dunn Chamber” [17] and “Insall Chamber” [18] fusive gradient that can direct cell migration. Te signals were designed to increase the throughput and achieve of chemoattractant were transmitted through interaction long-term observation of cell migration. Te gradient with heptahelical G protein-coupled receptors (GPCRs) generated in the Dunn Chamber was less susceptible expressed on cell surfaces. Once the chemoattractant to evaporation because source and sink chambers were interacts with its receptor on the neutrophil surface, they sealed. However, the gradient cannot be modifed once undergo cytoskeleton rearrangement, cell shape change solutions were loaded and the coverslip was secured, like and polarize. A “leading edge/pseudopod” is showed at Zigmond Chamber. Te Insall Chamber was designed the front, which pushes the cell forward and a “trailing for compatibility with thin coverslips for optimal optical end/uropod” is showed at the rear, which enables them properties and to allow use of high numerical aperture oil to migrate along a concentration gradient [13]. Tweedy immersion objectives. However, a large number of cells et al. [14] proposed an alternative possibility that the would be crushed during assembly and release intracel- cells themselves form a chemical gradient by degrading lular factors. Te released factors may have chemokinetic some ubiquitous local attractant. Tis self-generated gra- or chemotactic activities or synergistic efects in combi- dient drive cells migration away from their origin. Te nation with other factors. Terefore, the possibility that results showed that self-generated gradients can func- the present cell debris has unknown synergistic efects tion over arbitrarily large distances and operate over an can never be entirely discounted. In addition, traditional exceptionally wide range of ligand concentrations. Te devices were obsessed with many issues, such as large breakdown of attractant is absolutely required in self- number of repeated operations or consumption of rea- generated gradients. Breakdown mechanisms include gents. More importantly, these methods have difculties Zhao et al. J Transl Med (2020) 18:168 Page 3 of 19

to refect the biological characteristics of cells under microelectronics, biotechnology and chemical tech- physiological conditions compared with the complex nology. In 1990s, Manz et al. [49] frstly proposed the microenvironment in vivo and microscopic size of cells. concept of micro-total analysis system (µ-TAS) and In the recent years, highly controlled chemical gradi- verifed that electrophoretic separation in a µ-TAS was ents have been generated in microfuidic devices under more efciency than conventional methods. White- microscale with the development of microfuidic tech- side et al. proposed a fast template replication method nology [19]. Neutrophil migration and chemotaxis stud- to fabricate microfuidic device by polydimethlsilox- ies based on microfuidic technique have attracted broad ane (PDMS) in 1998, which signifcantly reduced the interest for its miniaturization and micro-environment manufacturing cost and time. Quake et al. [50] reported control. Compared with the traditional cell chemotaxis a large-scale integrated microfuidic device in 2002. devices, microfuidic chips have the unique advantages Briefy, microfuidic chip is a miniature biochemi- in cell migration studies as follows: (1) the micron-sized cal analysis instrument that integrates the experiment channel is equivalent to the size of cells, which is con- steps of biochemical sample pretreatment [51], particle venient for precise cell capture and manipulation. (2) manipulation [52], biochemical reaction [53], detec- Concentration, temperature, pH and other factors can tion and result analysis [54] into a device in a scale of be accurately controlled in the multi-dimensional net- tens to hundreds of microns, without sacrifcing their work channel structure. (3) Lower reagent consumption performance. Micro biochemical analysis units and is required compared with traditional methods. (4) High systems are built on the microfuidic chips by micro- throughput and high parallelism allow multiple sets of nano processing technology, enabling rapid detection samples to be processed simultaneously. (5) Modularity and analysis of biochemical samples such as organics, enhances the possibility of integration. Cell migration, inorganics, proteins, and nucleic acids. Te advantage cell capture, cell sorting and other units can be inte- of low cost, high-throughput, low sample consumption grated on a single chip. Microfuidic technology pro- and miniaturization and integration make microfuidic vides a new platform for neutrophils chemotaxis studies devices have been widely used in the felds of molecular under the high-controlled gradient conditions. However, biology [55], clinical diagnostics [56], and point-of-care it was found that the geometry of the microchannels had systems [57]. a great infuence on the cell migration process because Te fabrication processes of microfuidic device mechanical stress, force, and torque on the cell will be mainly depend on the selection of materials, the design amplifed as dimensions of the fow chamber are reduced of the structure, processing, and surface modifcation. and approach the diameter of the cell [20]. Terefore, the Monocrystalline silicon, quartz, glass, and high molecu- design of the microfuidic device is crucial for microfu- lar polymers are commonly used materials for microfu- idic-based neutrophil chemotaxis studies. idic devices fabrication. High-precision two-dimensional Various designs and strategies of neutrophil chemo- or three-dimensional structures can be formed on silicon taxis microfuidic devices have been developed over the by photolithography and etching. Quartz and glass have past decade. Te initial studies established stable and excellent optical properties and are easy to surface-mod- complex profle gradients in fow-based environment ify. Microstructure, such as metal microelectrodes, can [21–29]. To date, diferent materials and structures [4, also be processed on quartz or glass, which is similar to 30–43] have been used to reduce the impact of the fow silicon. Te polymers, such as polymethyl methacrylate shear force on neutrophils. In addition, some microfu- (PMMA), polycarbonate (PC), and PDMS, are very suit- idic devices [44–46], such as bioinspired microfuidic able for microfuidic devices. PMMA has great light assay (bMFA) [45, 46], were developed to study the kinet- transmission, impact resistance and durability. However, ics of leukocyte adhesion [47]. In the recent years, inte- it is difcult for surface modifcation and has poor bio- gration devices [26, 27] with the gradual maturity of the compatibility and reproducibility during the experiment. gradient generators were fabricated for clinical applica- PC is a tough thermoplastic resin with good light trans- tion [48]. mission, heat resistance, impact resistance, and oxidation Here, we classifed and discussed the structure and resistance. It has good mechanical properties in a suit- working principle of microfuidic devices based on the able temperature range, but it is easy to be corroded by microfuidic-based neutrophil chemotaxis studies in the organic chemicals. PDMS is the most widely used mate- past few years. rial for microfuidic devices. It has advantages of prefect light transmittance, strong chemical inertness, non-toxic, Microfuidic technology low cost, and easy for processing. It also can be modifed Microfuidic is a new biochemical small fuid volume using polyelectrolyte, surfactant coating layer, and phos- manipulating technology developed from the basis of pholipid bilayer membrane to meet the experimental Zhao et al. J Transl Med (2020) 18:168 Page 4 of 19

requirements. Terefore, PDMS is more suitable for interactions infuence human neutrophil migration and microfuidic devices. surface marker expression [23] (Fig. 1A). A stable con- Microfuidic experiments generally require microfu- centration gradient across the observation channel was idic devices with syringe pumps, microscopes, computer established when solutions introduced into the two inlets monitors, signal sources, oscilloscopes, etc. However, and mixed while traversing the serpentine channels. Tis with the growing maturity of microfuidic technology, it work explored interleukin-exposed neutrophil behavior become a new trend to integrate the external analytical in a concentration gradient of the bacterial chemoat- instruments onto the chip to meet the needs of portabil- tractant fMLP. Neutrophil migration against a bacterially ity and automation. derived fMLP, with and without pre-activation by inter- leukins, was evaluated in the presence and absence of Neutrophil chemotaxis microfuidic devices endothelial support cells, separately. Grigolato et al. [58] Te generation of chemical concentration gradients in proposed a microfuidic device that consisted of a gradi- the channels of microfuidic devices is vital in neutro- ent-generating serpentine network and fve observation phil chemotaxis studies. Based on the gradient genera- chambers for fully automated, quantitative assessment tion methods in the channels, microfuidic devices are of neutrophil chemotaxis. Te gradient was established mainly classifed as fow-based microfuidic devices and by injecting solutions at the two inlets of the gradient- fow-free microfuidic devices. Flow-based microfu- generating region of the device. Te epifuorescence idic devices require a constant fow to establish required microscope was used to assess the neutrophil migration. gradients, and fow-free microfuidic devices generate Tis device allowed the precise and reproducible deter- gradients based on the chemical difusion of the chem- mination of the optimal CXCL2 and CXCL8 concentra- oattractant. In addition, physical barriers, such as thin tions for mouse and human neutrophil chemotaxis. Te microgrooves [30–33], membrane [34–38], or gel [4, result showed that IL-4 receptor signaling has inhibition 33, 39–43], are used in fow-free microfuidic devices to in mouse and human neutrophils migration towards increase the fuidic resistance to control chemical dif- CXCL2 and CXCL8, respectively, and the inhibition is fusion. Microgrooves mean micro-sized narrow chan- time-dependent. nels that connected larger-sized microchannels flled Mao et al. [28] designed a T-shape microfuidic device with chemoattractant solution and medium. Te height to generate concentration gradients in the main chan- and width of microgrooves are about a few micrometers nel based on the difusion of the two fuids introduced which can reduce the convection of the liquid and gen- from two inlets. Based on this structure, Lee et al. [29] erate a stable gradient along the direction of the micro- further developed a converging–diverging microchan- grooves. Gel barrier is 3D network structure that allows nel to generate a near-constant concentration gradient difusion of molecules to generate a stable gradient in a in electroosmotic fow based on the enhanced difusive fow-free environment. mixing inside the converging–diverging microchannel section (Fig. 1B). Diferent concentration gradients were Flow‑based microfuidic devices established by adjusting the applied electric feld and Flow-based microfuidic devices generate chemical gra- microchannel’s geometry. Compared with the traditional dients based on the mixing of laminar fow in microfu- T-shape gradient generator, this device produced desired idic channels [21–29]. Stable gradients perpendicular to concentration gradients in a shorter mixing length the fuid direction are produced by controlled difusive (shorter by a factor of 2–3.5). mixing inside microchannels under conditions of low To generate gradient faster, a pressure balance zone Reynolds number. A variety of distribution of stable con- was usually used in fow-based devices [25, 26] (Fig. 1C). centration gradients are generated in microchannels by Tis design made the downstream gradient generation changing channel structures and fow rate. insensitive to the variation in solution volume of the Li et al. [22] developed a network of microfuidic chan- source inlet wells [25]. Terefore, the requirement of nels with multi-stage split recombination. Te device assay operation accuracy was reduced. Te structure of generated spatially and temporally linear concentra- the chips with a pressure balance zone was signifcantly tion gradients of IL-8 to study neutrophil chemotaxis. simplifed by using zigzag channels before and after the Human neutrophils migration in diferent IL-8 gradi- pressure balance zone. ent distribution was test in the microchannel. Te result A series of microfuidic networks devices were showed that neutrophils exhibit strong directional designed to generate complex concentration profles migration toward increasing concentrations of IL-8 in and diferent gradient profles fexibly [24, 59, 60]. As an linear gradients. Kim and colleagues proposed a similar example, Campbell et al. [24] designed a planar microfu- serpentine channel network to investigate how cell–cell idic networks device to produce concentration gradients Zhao et al. J Transl Med (2020) 18:168 Page 5 of 19

Fig. 1 Examples of fow-based microfuidic devices. A Multi-mixing, serpentine channel network microfuidic device. A stable concentration gradient was established across the observation channel after multi-stage split recombination in the serpentine channels. (Reprinted with permission from [23]. Copyright (2013) American Chemical Society.); B the converging–diverging microchannel to generate desired gradient in a short mixing length based on the enhanced difusion in the channel. Reprinted from Ref. [29], Copyright (2005), with permission from Elsevier. C the microfuidic device with a pressure balance zone unit for stabling the fows from diferent source inlet wells to generate concentration gradients in the gradient channel. (Figure reproduced from Ref. [25]); D the planner microfuidic networks device to establish concentration gradients with any given monotonic function shape. Republished from Ref. [24], copyright 2007, with permission of Royal Society of Chemistry; permission conveyed through Copyright Clearance Center, Inc

with the shape of any given monotonic function (Fig. 1D). by laterally combining the constituent profles generated Te source solutions were repeatedly split and mixed in simple Y-shape or Ψ-shape generator. in a series of k = log2(N−1) stages, and then were com- Te separation of spatial gradients and temporal gra- bined to create a single stream with the desired shape of dients is a problem in neutrophil chemotaxis studies concentration profle. Tree networks that generated an because the controversy that either the spatial sensing exponential concentration profle, a linear profle, and a mechanism or the temporal sensing mechanism is prin- profle with a shape of two branches of a parabola with ciple response for chemical gradient sensing still exists k = 4 and N = 17 were built and tested. Meanwhile, Hat- [61]. Neutrophils were exposed under both spatial and tori et al. [59] reported a serial dilution microfuidic temporal concentration changes when they moved in network with a high fuidic-resistance ratio to generate response to a heterogeneous chemical environment. linear concentration profles and logarithmic concentra- Te separation of spatial and temporal stimuli will pro- tion profles spanning 3 and 6 orders of magnitude. Zhou mote progress towards understanding the mechanisms et al. [60] designed a concentration gradient device that of cell chemotaxis. Aranyosi et al. [62] built a neutro- established complex profles of chemical concentration phil treadmill system that held a moving neutrophil at Zhao et al. J Transl Med (2020) 18:168 Page 6 of 19

a specifed, unchanging location in a chemical gradient, Flow‑free microfuidic devices which decoupled the spatial and temporal gradients To overcome disadvantages of fow-based microfuidic around moving cells. Te fow within the device was devices, fow-free microfuidic devices with chemical gravity driven with reservoirs of chemokine and bufer concentration gradients in fow-free environment based and the chemokine difusion generated concentration on the chemical difusion were developed to reduce the gradient. Te location of neutrophil within the gradi- shear forces on neutrophils. For instance, Walsh et al. ent was changed continuously and precisely to com- [63] designed a paper-fuidic device to investigate cell pletely restrain the temporal stimulus on the moving chemotaxis (Fig. 2). All the forces necessary to build neutrophil. Tis study proved that temporal gradients competing gradients were provided by the wicking action are necessary for the directional persistence of human of paper because when two fuids traversing paper from neutrophils during chemotaxis. diferent direction came in contact, difusion between Flow-based microfuidic devices create gradients fuids established a semi-stable gradient on a time scale through the mixing of laminar fow in microchannels. suitable for cell chemotaxis. Results showed a stable High-controlled chemical gradient in various distri- gradient of chemokine was maintained for 20 min, with bution profles can be formed in the channel just by a sharp gradient taking place over 2 mm. Tus, cells changing the structure of the devices or the fow rate of directly exposed to the gradient were easily to response the liquid. Te gradient establishing-time was shorten the change in chemokine concentration. Te analysis of to a few milliseconds. However, most of the fow-based the gradient on paper was shown in Fig. 2b. It can be seen microfuidic devices require peripheral chemical perfu- that there appeared to be some relaxing of the gradient sion and neutrophils are exposed to the constant shear over the course of 20 min, however, the efect was very fow since the gradient are established depending on minimal. Te gradient stretched over 3 mm at the time of the difusion across laminar streams. Te direction zero minutes, and stretch out only on a scale of microm- of fuid fow (the direction of neutrophils movement eters over 20 min. Tus, this paper device provided a caused by shear forces) is along the channel direction, quasi-steady state gradient over the 3 mm gradient area and the direction of neutrophil chemotaxis is perpen- for the length of the experiment. Te proof-of-concept dicular to the channel direction. Tus, the movement experiment showed signifcant directed migration of of neutrophils are afected by shear forces. Most of the cells to the chemokine gradient over the control condi- autocrine/paracrine factors secreted by neutrophils tion. Weckmann et al. [64] used a chemotaxis 3D µ-slide will be wash away, which may infuence the chemotaxis (IBIDI) to monitor and quantify neutrophil chemotaxis behavior of neutrophils. In addition, the chemical gra- through combining with simple video microscopic equip- dient distribution will change along the channel direc- ment and highly standardized tracking. A stable gradi- tion, which will also afect the neutrophil chemotaxis ent was generated throughout the channel connected to studies. Te stability of the gradient profles is limited two chemoattractant-flled chambers. Typical migration by the fow rate stability. For this method, sample and profles of the chemoattactants IL-8, fMLP, and LTB­ 4 reagent consumption cannot be ignored when high was identifed by using this simplifed migration analy- fow rates are required. sis (SiMA) system. To simulate the interstitial spaces and

Fig. 2 Examples of fow-free microfuidic devices. a Gradient generation in paper device; b analysis of the paper gradient using fuorescent dextran. False color images of the gradient at 0 and 20 min (left). Normalized pixel intensity analysis over the image from left to right for the gradient at 0 and 20 min (right). Reprinted with permission from [63]. Copyright (2015) American Chemical Society Zhao et al. J Transl Med (2020) 18:168 Page 7 of 19

chemoattractant gradients, Boneschansker et al. [65] microgrooves-based devices, membranes-based devices, used microfuidic maze to analyze the migration process and gel-based devices. For membranes-based devices and of neutrophils. Te mazes mimicked the confnement gel-based devices, membranes and gel are used as physi- found in interstitial spaces by restricting the migration cal barriers to control chemical difusion and separate the of neutrophils through channels with 10 × 10 µm cross- gradient generation chamber and the fuid fow chamber. section. Te mazes were connected to the reservoir Terefore, the shearing efect in the gradient generation that generated a chemokine gradient, which persisted chamber is minimum. for more than 3 h. Cells in the cell-loading chamber migrated towards the source of chemoattractant through Microgrooves‑based devices mazes (550 µm in length and 350 µm in width). Te Microgrooves between channels are a distinct class of result showed diferent migration patterns of neutrophils fow-free microfuidic gradient generators [30–33, 66]. for diferent chemoattractants. For ­LTB4 and fMLP gra- As shown in Fig. 3A, the “Ladder Chamber”, proposed by dients, neutrophils had highly directional migration pat- Saadi et al. [32], generated the steady state gradients in terns and moved towards the source of chemoattractant. the microgrooves which connected two parallel channels However, neutrophils had a low-directionality migration in fow-free environment. Te height and width of micro- pattern and dispersed within mazes, for C5a and IL-8 grooves was 3–10 µm and 10 µm respectively, which gradients. were signifcantly smaller than the two parallel channels Flow-free microfuidic devices overcome the disadvan- (100 µm in height and > 1 µm in width). Te fuidic resist- tage of the fow-based microfuidic devices and gener- ance across the microgrooves was high compared to the ate gradients by free difusion of chemicals in fow-free parallel channels, confning bulk fuid fow in the paral- environment. Tus, they are less dependent on exter- lel channels, and minimizing the likelihood of cross-fow nal controls and the shear forces efect on cells is mini- through the microgrooves. Terefore, difusion would be mized. However, they tend to generate shallow gradients the predominant mode of transport across the micro- and are diferent to create complex gradient profles. grooves. Molecules difused across the microgrooves With the development of the fow-free microfuidic and generated a gradient when they introduced in one devices, they were mainly classifed into three types, of the parallel channels and bufer in the other. When

Fig. 3 Examples of microgrooves-based devices. A Two-channels microgrooves-based microfuidic device. The two high channels connected by a series of thin microgrooves are loaded with diferent chemicals to generate gradients in microgrooves. (Figure reproduced from Ref. [32]); B Illustration of the D­ 2-Chip. The gradient channel is connected to the source channel and the sink channel by 3 µm microgrooves defned as the docking structures. The source inlets are loaded with equal volumes of medium with and without chemoattractant, and a stable linear chemoattractant gradient is generated in the middle gradient channel based on the difusion of the chemoattractant solution and the medium. (Figure reproduced from Ref. [70]); C Schematic diagram of the novel microfuidic competitive chemotaxis-chip(µC3). Healthy neutrophils(blue), super-low dose (1 ng/mL) of LPS (red neutrophil) and high dose (100 ng/mL) of LPS (orange neutrophil) show diferent migration under the environment with and without chemoattractant gradients. The gradient was formed within the migration channels from the chemoattractant reservoir to the central loading channel and stayed stable for a long time. (Figure reproduced from Ref. [71]) Zhao et al. J Transl Med (2020) 18:168 Page 8 of 19

the system reached steady state, a linear concentration migrated toward the primary end target signal in higher was established. Neutrophil chemotaxis was successfully percentages than toward the secondary intermediary observed in soluble IL-8 gradient. Irimia et al. [30] devel- signal by using fMLP to model an end target chemoat- oped a similar microfuidic device for precise, passive tractant and LTB­ 4 to model an intermediary chemoat- balancing of fow by contacting two fuid streams before tractant. However, the adding of super-low dose LPS splitting them again and a convection-free, stationary signifcantly increased the percentage of cells migrate linear or moving steep gradients of chemoattractant was toward the intermediary signal. In addition, the result established in an array of microgrooves. showed that super-low and high levels of LPS stimula- Another type of gradient generators with three main tion both induced spontaneous neutrophil migration in channels are common devices for neutrophil chemotaxis the absence of chemoattractant gradients. studies [67–70]. Te parallel side channels are usually set Tere were some other microgrooves-based neu- as the source and sink channel, which are connected to trophil chemotaxis microfuidic devices. For exam- the middle gradient channel through barrier structure ple, Hamza et al. [72] designed a U-shape microfuidic such as microgroove channels. Te concentration gradi- device to simulate the biochemical and mechanical ents are generated in the middle gradient channel instead confnement condition at sites of injury in tissues and of in the microgrooves. Te open-chamber “Microjets” study human neutrophil chemotaxis in response to device, which was designed by Keenan et al. [67], estab- chemoattractant gradients inside channels. Te device lished a stable and reproducible gradient in the middle was composed of a main channel to load with the che- chamber within 4 min through fast mixing of jets emit- moattractant solution and neutrophils and a series of ted from the source and the sink channel. Te “Microjets U-shape channels to generate gradients. Te chemoat- Device” was used to develop a method for studying gra- tractant gradient was established in the U-shape chan- dient-induced homologous and heterologous neutrophil nels by the difusion between the U-shape channels and desensitization [68]. Recently, Yang et al. [70] developed a the main channel. Te concentration was highest at the ­D2-chip which was composed of a middle gradient chan- tip of the U-shape channels and decrease along the two nel, an outer source channel, and an outer sink channel arms of the U-shape. Te result showed that more than (Fig. 3B). Te source/sink channel were connected to the 90% of neutrophils reversed the direction and migrated middle channel by thin barrier channels. Neutrophils persistently after initially moving towards to the higher were loaded into the middle channel and were pushed chemoattractant concentration, and the migration dis- towards the channel walls under the pressure diference tances away from chemoattractant sources (retrotaxis) between the middle channel and the outer channel and were longer than 1000 µm. aligned along the two sides of the middle channel due to Microgrooves-based devices use both convective the barrier channels. When the chemoattractant solution and difusive transport to establish a gradient in the and medium reached the source and sink channel, the cell chamber with no direct shear stresses imposing on gradient was generated in the middle gradient channel the cells [73]. Te bulk fow is signifcantly reduced in within 1 min. Te ­D2-chip tested neutrophil chemotaxis the gradient forming channel when the fow rates in and the memory efect. the parallel sample and bufer channels are identical. Neutrophil migration process is complex and guided Tus, the efect of shear forces on cells are very small by multi-chemoattractants released from injured tis- in microgrooves-based devices. Complex gradient pro- sues. Boribong et al. [71] designed a microfuidic com- fles can be generated by juxtaposing diferent designs petitive chemotaxis-chip (µC3) to measure neutrophil within a single gradient-generating region [74]. Lin- migration in a competitive chemoattractant environ- ear gradients are produced when the width of micro- ment completing chemoattractant gradients in the grooves at opposite ends are equal, while nonlinear central channel where cells were exposed (Fig. 3C). gradients are produced when the width are unequal. Te µC3 chip was a pump-free stand-alone microfu- Complex gradient profles can be established by a serial idic dual gradient device, which consisted of chemoat- combination of diferent microgrooves designs within tractant reservoirs for end target chemoattractant (i), on gradient generation region. Te main limitations of chemoattractant reservoirs for intermediary chemoat- these devices are the requirement of more complicated tractant (ii), central loading channel (iii), migration multi-depth channel fabrication and the limited gradi- channels (iv), and ladder maze (v). A linear gradient ent channel height required for gradient generation. In was formed along the length of the migration channels addition, this kind of devices rely on precisely match- within 15 min, and the completing chemoattractant ing sample and bufer fow rates. Terefore, they are gradients were generated in the central channel where very sensitivity to the mechanical disturbance from the cells were exposed. It was found that naïve neutrophils microfuidic system [38]. Zhao et al. J Transl Med (2020) 18:168 Page 9 of 19

Membranes‑based devices fow-free concentration gradient generator (CGG) micro- Porous membranes were also used in fow-free micro- fuidic device that included an upper PDMS layer (sample fuidic devices to bufer fuid shear forces and permit the and bufer channels), a bottom layer (a gradient genera- difusive exchanges of soluble factors between chambers. tion microchamber), and a semipermeable membrane Tus, porous membranes enabled a vertical alignment sandwiched between two layers (Fig. 4). Te CGG device of the fow and cell chambers, which was convenient not only maintained a stable concentration through the for cell culture and imaging with inverted microscopes free difusion across the membrane but also separated [34–38]. Kim et al. and Vandersarl et al. reported micro- the concentration gradient in the lower layer and the fow fuidic devices that used porous membranes to separate of reagent sample and bufer in the upper layer. By adjust- the generated-gradient from the cell chamber [34, 35]. ing the geometries of channels, concentration gradients Te fow and cell chambers were allowed aligning in ver- with diferent shapes were generated in the bottom layer. tical, which was convenient for cell culture and imaging Te fow-free environment where concentration gradi- with inverted microscopes. Chung et al. [37] developed ents generated eliminated the undesirable fow stimula- a chemotaxis system in which two chambers were sepa- tion on neutrophils. rated by a thin (15 nm), transparent and nano-porous sil- Membranes-based devices couple the channel to the icon membrane that provided efectively no resistance to reservoirs through a semipermeable membrane, which molecular difusion between the two chambers to create eliminates bulk fow while allowing molecular difusion. fow-free chambers in the microfuidic system. Te mem- Membranes enabled a vertical alignment of the fow and branes also had excellent optical properties for phase and cell chambers, which was convenient for cell culture and fuorescence microscopy. Te upper chamber provided imaging with inverted microscopes. Tis approach is fow-generated gradient and the lower chamber provided efective in eliminating fuid disturbances in the gradi- a shear-free environment for cell observation. Shear ent forming channel. Diferent shapes of the gradients reduction of more than fve orders of magnitude was can be established by changing the geometries of the predict by considering analytical and computational fow microchannels and chambers, but the gradient profles models that account for membrane and chamber geome- also diminish over an extended period of time [38]. Te try. Migrating neutrophils were exposed in a chemotactic gradient establishment time is long, which depends on gradient or fuorescent without any infuence from fow. the geometry of the device as well as the size of the mol- Similarly, Zhou et al. [38] presented a membrane-based ecules [73].

Fig. 4 The device confguration of the membrane-based concentration gradient generator. The solid arrows denoted fuidic path, and the dotted arrows denoted difusion paths. The concentration gradient in the gradient forming channel was controlled to be stable by the running sample and bufer solutions in the respective overlaying channels. Reprinted from Ref. [38], Copyright (2013), with permission from Elsevier Zhao et al. J Transl Med (2020) 18:168 Page 10 of 19

Gel‑based devices that temporal sensing mechanisms controlled prolonged Hydrogel could provide stable difusion of chemoattract- responses to these ligands. ant molecules. Te difusion rate of molecular in gel is Wu et al. [41] developed a versatile hydrogel-based slow due to the 3D cross-linked network of gel, which microfuidic platform to mimic in vivo neutrophil promotes to establish a chemical gradient with long-term transendothelial migration (TEM) process (Fig. 5B). stabilization [41]. Terefore, the combination of hydro- Hydrogel provided mechanical support for the growth of gel and microfuidic devices is a very common method to an endothelial cell layer in perpendicular direction and generate predictable, reproducible, and long-term stable highly stable chemical gradients. Te results showed that chemical gradients with high spatiotemporal resolution the number of neutrophils migrating across the endothe- [4, 33, 39–43]. lial cell layer had important relationship with the chem- Cheng et al. [75] developed a hydrogel-based micro- oattractant concentration and the spatial profle of the fuidic device to generate a stable and long-term linear chemical gradient. chemical concentration gradient with no through fow in Gel-based devices eliminate fow disturbance in the a microfuidic channel. Tree parallel microfuidic chan- gradient forming channel through hydrogel, which pro- nels were patterned on a thin piece of agarose gel. Fluid vide sample molecules difuse. Tey are able to main- with a constant chemical concentration fowed in one tain temporally non-diminishing gradient profles with outer channel (the source channel). Te fuid with a black constant replenishment of sample and bufer. Complex bufer fowed in the other outer channel (sink channel). concentration gradients profles could be generated by Te chemicals difused across the channels, and a linear design diferent gradient forming channel shape. How- chemical concentration gradient was generated in the ever, this method needs long generating times (about center channel when the system reached steady state. a few hours) for the concentration gradients due to the Te chemotactic response of a suspension cell line— slow molecular difusion in hydrogel [38]. In addition, the Escherichia coil RP437 and an adherent cell line—difer- optical transparency of hydrogels is relatively poor com- entiated human promyelocytic leukemia cell line, HL-60 pared to PDMS or glass, which hinders phase-contrast was monitored successfully using this device. Ahmed microscopy [77]. Further improvement and innovation et al. [76] characterized the hydrogel gradient genera- are required to enable more fexible control of gradient tors by confocal microscopy and numerical simulation generation. and apply it to chemotaxis experiments with Escherichia colt in both linear and nonlinear gradients. Te observed Integrated neutrophil chemotaxis devices cell distribution along the gradients and the established Combined with cell culture unit mathematical model showed very good agreement. In most of the single-function microfuidic neutrophil Abhyankar et al. [40] proposed a method that provided chemotaxis devices mentioned above, cells were injected linear and non-linear soluble factor gradients within a into the microchannels because long-term cell culture 3D gel matrix by combining variable channel geometries in microchannels is challenging due to shear sensitivity, with the principle of infnite sources and sinks. Te con- especially for sensitive cells [78]. With the development centration profles were maintained for up to 10 days, of the shear-free environment, some researchers aimed and the temporally evolving and long-lasting gradi- to combine the gradient generation unit and the cell cul- ents were applied to study the chemotactic responses ture unit on the same chip [31, 36, 78–81]. Joanne et al. of human neutrophils and the invasion of metastatic rat [79] proposed a microfuidic-based turning-assay chip mammary adenocarcinoma cells (MtLN3) within 3D col- that consisted of gradient generating networks and cell lagen matrices. seeding channels. Te device generated precise and com- To eliminate the inherent coupling of the fuid fow plex composite gradients to mimic the conditions the and chemical concentration gradients in 3D microfu- growth cones realistically counter in vivo and study how idic chemotaxis device (µFCD), Haessler et al. [42] pre- neuronal growth cones migrate in response to complex sented an agarose-based 3D µFCD to decouple these combinatorial gradients of diverse external cues. Kim two important parameters by using an agarose gel wall. et al. [36] designed a microfuidic device for cell culture It provided the adequate physical barrier for convec- and chemotaxis studies. Vertical membranes formed by tive fuid fow and protein difusion at the same time to in situ fabrication were used to avoid fuid fow inside the separate the fow control channels from the cell compart- cell observation chamber. Neutrophils were introduced ment (Fig. 5A). Petrie et al. [4] used the agarose-based 3D in the observation chamber and incubated for 30 min, µFCD to study the relationship of the concentration of then the mixture of IL-8 and fMLP was introduced in the intermediate chemokines (CCL19 and CXCL12) and the source chamber. Successful migration of neutrophils up migration of dendritic cells or neutrophils. Tey found to the concentration gradient of IL-8 was exhibited by Zhao et al. J Transl Med (2020) 18:168 Page 11 of 19

Fig. 5 Examples of gel-based devices. A The device schematics of the 3D microfuidic chemotaxis device. The device consisted of four three-channel units. Cells and collagen were injected into the center channel together. The chemical gradient was generated in the center channel by introducing media containing diferent concentration chemoattractant through the two side channels. (Figure reproduced from Ref. [42]); B Schematic of gel-based neutrophil TEM microfuidic device. Endothelial cells are cultured on the side wall of the collagen gel, and the chemical gradients are developed by placing the chemoattractant solution or medium on the side channels. Neutrophils will across the endothelial cell layer and move towards the chemoattractant source as the black arrow. Reprinted from Ref. [41], Copyright (2015), with permission from The Royal Society of Chemistry

experiment. Over 91.7% of neutrophils migrated toward device hanged self-supported at a distance of ~ 250 µm the higher concentration, and the longest distance of the above the cell culture surface by simply inserting into a neutrophils travelled in 25 min was 162.5 µm toward the standard 6-well plate. Te microfows with stable and source. Te average rate in the x direction and in the y quantifable concentration gradients generated by the direction was 3.44 µm/min and 0.25 µm/min, respec- device moved the integrated track-etched porous mem- tively. Zhang et al. fabricated a dual-functional micro- brane, then entered into the cell culture well. Tis device fuidic chip based on rolling circle amplifcation for demonstrated long-term stability of gradient profles over cell culture and online IL-8 detection, and successfully a large area of approximately 25 mm square and realized applied to analyze the secreted IL-8 in endothelial cells the function of analysis the chemotaxis of a large number [31]. To study the dynamics of neutrophil chemotaxis of cells through direct visualization and automated image under competing chemoattractant gradient, Kim et al. analysis. An fMLP gradient was applied to a large popula- [80] proposed a microfuidic platform, which established tion of HL-60 cells (a neutrophil cell line) by using this a stable and dynamic gradient of chemoattractant across device. From the total 282 cells observed in the gradient, the cell culture chamber (Fig. 6A). Human neutrophils over 74% of the cells (208 cells) moving towards the gra- were exposed in competing gradients of four diferent dient, which was the result of quantifcation by the chem- chemoattractants (leukotriene B4, chemokine C-X-C otactic response with an automated tracking algorithm. motif ligands 2 and 8, and fMLP). Over 60% of neutro- Te combination of chemotaxis devices and cell culture phils moved toward the stronger signal and the results unit reduce the interruption of cell–cell communication showed a hierarchy among these chemoattractants of caused by the fuid fow and convection. Tis method leukotriene B4, chemokine C-X-C motif ligands 2 and will provide more convenient ways to investigate more 8, and fMLP. Sip et al. [78] reported a microfuidic tran- cell migration behaviors, such as aggregation and disper- swell insert which was compatible with conventional cell sion under determinable concentration gradients while cultures and with tissue explant cultures (Fig. 6B). Te retaining paracrine signaling [36]. Zhao et al. J Transl Med (2020) 18:168 Page 12 of 19

Fig. 6 Examples of integrated neutrophil chemotaxis devices combined with cell culture units. A Microfuidic device that combines multiple networks and cell culture chamber where the stable gradient is generated. Reprinted with permission from Ref. [80], copyright (2012) American Chemical Society; B schematic diagram of microfuidic transwell insert. The device hangs self-supporting in a 6-well. The concentration gradient is generated in the external fuid space (color map) between the track-etched membrane (green line) and the cell culture surface that is approximately 250 µm in height. Reprinted from Ref. [78], Copyright (2014), with permission from The Royal Society of Chemistry, permission conveyed through Copyright Clearance Center, Inc

Combined with isolation unit improvement and application [87–92, 94]. Te device Most mentioned neutrophil chemotaxis experiments consisted of a whole blood loading chamber (WBLC) require purifed neutrophil samples. However, the pro- and a series of focal chemotaxis chambers (FCCs) flled cess of purifcation needs a long time and may cause with chemoattractant. Te generation of the chemotac- damage to neutrophils. To remove the purifcation pro- tic gradients was based on the difusion from the FCCs cess such as ultracentrifugation, Mathias et al. proposed in the absence of convection [87]. Neutrophils migrated a method that stimulated whole blood from umbilical directly from the blood droplet, through small chan- cord blood (n = 6) and healthy control subjects (n = 6) for nels, towards the source of chemoattractant due to the 24 h with 100ngs/mL of LPS and subsequently isolated ability to deform actively during chemotaxis through CD66b+ with microfuidic technology [82, 83]. Te result microscale channels that block the advance of other showed that LPS stimulated whole blood from umbili- blood cells [91]. Jones et al. [87] showed that phlogis- cal cord blood (UCB) demonstrated signifcant difer- tic and nonphlogistic cell recruitment were distinguish ences in both ex vivo cytokine production and PMN gene by integrating an elastase assay into the FCCs. Te che- expression. moattractant gradient along the migration channel to To further simplify the experimental process and the cell loading chamber was established within 15 min reduce the damage caused to the neutrophils by the puri- and decreased by less than 10% at 6 h. Monocytes and fcation process, some researchers aim to integrate sepa- neutrophils induced by ­LTB4 showed signifcant difer- rating units and chemotaxis study units in the same chip ence in dynamics. Neutrophils moved through chan- [26, 27, 48, 84–94]. Terefore, the integrated microfuidic nels at ~ 18 µm/min and reached the FCC in less than devices for neutrophil chemotaxis analysis directly from 30 min, while the velocity of monocytes was about whole blood have been becoming a growing trend. 5 µm/min and the time of reached the FCC was more As one example in this direction, Irimia et al. designed than 90 min. Te level of elastase produced in real time a donut-shaped structure (Fig. 7a) to measure neutro- by monocytes accumulating in the nano-liter chambers phil chemotaxis from whole blood by designing imple- was measured to distinguish between phlogistic and mented mechanical flters with right angle to selectively nonphlogistic recruitment of monocytes. An infection- block the advance of red cells and made a series of infammation-on-a-chip-model was developed. Te Zhao et al. J Transl Med (2020) 18:168 Page 13 of 19

Fig. 7 Examples of integrated neutrophil chemotaxis devices combined with isolation units. a The donut-shape microfuidic device. Chemoattractant is introduced into the focal chemotaxis chambers (FCCs), and a gradient is generated along the migration channels towards the FCCs. The whole blood is put into the whole blood loading chamber (WBLC). The red blood cells are prevented by the RBC fltration comb and only neutrophils can migrate out of the whole blood and accumulate in the FCCs. This is adapted from Jones, C. N., Hoang, A. N., Dimisko, L., Hamza, B., Martel, J., Irimia, D. Microfuidic Platform for Measuring Neutrophil Chemotaxis from Unprocessed Whole Blood. J. Vis. Exp. (88), e51215, https​:// doi.org/10.3791/51215 ​(2014). b The all-on-chip device for neutrophil chemotaxis analysis. The device consists of the cell docking barrier channel (4 µm) and the gradient generation channel (60 µm). This is adapted from Yang, K., Wu, J., Zhu, L., Liu, Y., Zhang, M., Lin, F. An All-on-chip Method for Rapid Neutrophil Chemotaxis Analysis Directly from a Drop of Blood. J. Vis. Exp. (124), e55615, https​://doi.org/10.3791/55615​ (2017); c Overview of the KOALA platform. Compared with the traditional methods like transwell assay that take many hours to complete all the processes, this platform signifcantly reduces the experiment time. (Figure reproduced from Ref. [43])

dynamic equilibrium between migration, reversed- Based on the pressure balance zone, Yang et al. [26] migration, and trapping processes determine the developed an all-on-chip method for integrating the optimal number of neutrophils at a site and these neu- magnetic negative purifcation of neutrophils and chem- trophils are continuously refreshed and responsive otaxis assay from small blood volume samples (Fig. 7b). to the number of microbes [90]. Tey also found the Tis device achieved a rapid sample-to-result neutrophil important diferences among migration counts, veloc- chemotaxis test in 25 min. At least 25% of the neutrophils ity, and directionality among neutrophils from 2 com- from input whole blood sample efectively entered the mon mouse strains, rats, and humans by using this docking structure and the purity of neutrophils was high donut-shaped [91]. by on-chip Giemsa staining. Gradient were generated Zhao et al. J Transl Med (2020) 18:168 Page 14 of 19

based on the continuous laminar fow chemical mixing, device achieved increased functionality and simplicity of and the fows were driven by the pressure diference from operation. Moussavi-Harami et al. [86] reported a micro- the diferent levels of inlet and outlet solution. Te chem- fuidic device for simultaneous analysis of NETs and ROS. ical gradient was established within a few minutes in the P-selectins were coated on the bottom of the channel to microfuidic channel and kept stable for at least 1 h. Tis capture and purify neutrophils, and it achieved purify device was used to test neutrophil chemotaxis in COPD primary human neutrophil in less than 10 min from a few sputum and the result showed a strong cell migration to microliters of whole blood. Te device showed the ability the COPD sputum gradient. to distinct the neutrophil subsets (including POS produc- Moussavi-Harami et al. [85] reported a “PI” channel for tion and NET formation) in diferent stimulants/inhibi- studying chemotaxis through a 3-D environment in the tors. Te sample was still efectively used after storing for presents of dual chemotactic gradients and successfully up to 8 h in the device. performed neutrophil chemotaxis experiments with the Tay et al. [27] reported an integrated microfuidic whole blood sample in 2015. Te PI channel included a device for single-step neutrophil sorting and chemotaxis chemoattractant containing gel which was perpendicu- study using a small blood volume. Te purifcation of neu- lar to a cell placement channel. Te gel played as the trophils from whole blood was based on the biomimetic 3D platform for cell migration from the cell placement cell margination and afnity-based capture. Te separa- channel. Te gradients were generated in the gel to more tion efciency of leukocyte (CD45+) was more than 80% closely mimic complex in vivo gradients. Moreover, the and the device achieved 12-fold leukocyte enrichment basement membrane-like substrate Matrigel was used in at the side outlets. Te captured neutrophils were then the gel region to maintain a stable gradient, without any exposed in the difusion-based chemotactic gradients manipulation of the device once the system settled. Te environment to initiate chemotaxis. Te result showed migration of the neutrophil-like cell line PLB-985 in gra- signifcant reduce of TNF-α and glucose-treated neu- dients of fMLP was tested using this device. Te PLB-985 trophils migration toward fMLP (≈ 45%) compared with cells were observed to migrate in the presence of fMLP healthy neutrophils migration (≈ 66.2%), while exhibited or ­LTB4 gradient with a velocity of 7.73 µm/min in single a signifcant decrease in cell migration. A decrease in fMLP gradient and 8.01 µm/min in double fMLP gradi- chemotaxis velocity was also observed in TNF-α treated ent, while no migration was observed in control setups neutrophils (≈ 5.22 ± 0.37 µm min − 1) when compared −1 where no chemoattractant in the device. Furthermore, to healthy neutrophils (≈ 6.27 ± 0.59 µm min ). In addi- this device was used with heparinized whole blood and tion, distinct neutrophil chemotaxis suppression was neutrophils were observed to migrate into the gel over observed in vitro infamed blood sample within few min- 2.5 h with an average velocity of 3.7 µm/min without any utes. Terefore, the method can be used in rapid assess- neutrophil purifcation or capture steps. ment of neutrophil functions. Functional modifcation of selectins or antibodies is also a common method to purify neutrophil in microfu- Combined with the analysis unit idic device. Agrawal et al. [84] frstly fabricated a micro- Te real-time diagnosis and portability of instruments fuidic chip for neutrophil chemotaxis studies using have become more and more important as medi- P-selectin as the substrates for neutrophil isolation in cal health is gradually valued by people. Terefore, it is 2008. Sachmann et al. [43] reported a KOALA platform emerging research fled to make real-time detection and that consisted of a lid and a base, and the device achieved analysis systems by combing the microfuidic device with neutrophil purifcation and chemotaxis on-chip within the analysis unit. Wu et al. [95] developed a compact USB minutes in 2012 (Fig. 7c). It only required nanoliters of microscope-based microfuidic chemotaxis analysis sys- whole blood as the sample and a micropipette to oper- tem (UMCAS) which integrated microfuidic devices, ate. Neutrophils were specifcally captured on the poly- live cell imaging, environmental control and data analy- styrene surface functionalized by P-selectin. Te device sis (Fig. 8A). It provided a platform for rapid microfuidic achieved ~ 80% capture efciency of human primary neu- neutrophil chemotaxis experiments with real-time data trophils. Te lid was used to house the reagents required analysis and wireless remote data monitoring. to generate the gradient of chemoattractant. After cap- Mobile sensing based on the integration of microfu- turing, the lid was placed onto the base, thereby allow- idic device and smartphone (MS­ 2) technology is a fast ing the chemoattractant in the lid to controllably difuse developing fled in the recent years. It has been used in into the microchannel to form a gradient of chemoat- a wide range of application, such as biochemical detec- tractant. By using this method, the difculty of perform- tion and disease diagnosis [96]. Compared with the ing challenging experiments was signifcantly reduced. traditional platforms, MS­ 2 technology provides signif- Compared with the traditional standard technology, this cant advantages in terms of test speed and control, low Zhao et al. J Transl Med (2020) 18:168 Page 15 of 19

Fig. 8 Examples of integrated neutrophil chemotaxis devices combined with analysis units. A The USB microscope-based microfuidic chemotaxis system (UMCAS). This system contains microfuidic device, live cell imaging, control and analysis units. (Figure reproduced from Ref. [95]); B components and operation fow of the Mkit. Reprinted from Ref. [48], Copyright (2017), with permission from Elsevier

cost, ease-of-operation and data management. Recently, chemotaxis and assess the severity and prognosis of sep- Yang et al. developed a ­MS2-based cell functional assay sis (Fig. 9a). Two main channels were flled with poly- for testing cell migration (the Mkit) based on their pre- morphonuclear neutrophils (PMNs) and a collagen gel vious work [26, 48] (Fig. 8B). Tis system combined that was spiked with LPS, respectively. Te collagen gel microfuidic chip, a smartphone-based imaging platform, containing the LPS was used to generate a concentra- the phone apps for image capturing and data analysis, tion gradient in microgrooves and reduce the shear stress and a set of reagent and accessories for performing the acting on the PMNs. PMNs moved in the microgrooves cell migration assay together. With the optical compo- towards gel channel driven by the LPS concentration gra- nents, the system achieved ~ 3 µm resolution which was dient. Neutrophils of 32 sepsis patients were divided into adequate for imaging of many types of cells. Cells in the three groups according to the seriousness, and 12 healthy images could be identifed by the image processing and individuals served as controls. Results showed that neu- data analysis app, and the cell migration distance along trophil chemotaxis was signifcantly decreased following the gradient direction could be calculated as a measure of the seriousness of sepsis. chemotaxis. Tis system successfully measured chemot- Raymond et al. [92] provided the frst description of axis from purifed neutrophil, cancer cells. Furthermore, neutrophil chemotaxis and transcriptomics from whole neutrophil chemotaxis was tested from whole blood sam- blood of human term and preterm neonates, as well as ple and clinical samples from chronic obstructive pulmo- young adults using the donut-shape microfuidic device nary disease patient. (Fig. 9b). Ex vivo whole blood chemotaxis was measured Tis type of system that integrated detection and anal- to the fMLP for term neonates (n = 20), preterm neonates ysis units greatly reduces the requirement of operators. (n = 20), and young adults (n = 15). Neutrophils from However, it is still in the initial stage of research. A series preterm neonates migrated in fewer numbers compared of problems needed to be solved, such as low resolution, with term neonates (preterm 12.3%, term 30.5%), and poor stability, and few analytical functions. at a reduced velocity compared with young adults (pre- term 10.1 µm/min, adult 12.7 µm/min). Ex vivo sponta- Applications neous neutrophil migration, neutrophil transcriptomics, With the gradual maturity of the technology, microfuidic and cytokine production in the presence and absence of neutrophil chemotaxis devices have more and more clini- LPS were measured directly from whole blood of adults, cal application. term neonates, and preterm neonates to understand For example, combined microgroove structure and the response of human neonatal neutrophils to toll-like hydrogel, Lu et al. [33] fabricated a channel-micro- receptor (TLR4) stimulation [93]. Te results showed grooves-channel microfuidic chip to measure neutrophil signifcantly fewer spontaneously migration neutrophils Zhao et al. J Transl Med (2020) 18:168 Page 16 of 19

Fig. 9 Examples of microfuidic neutrophil chemotaxis devices in clinical applications. a The overview of the gel-based neutrophil chemotaxis device. The gel channel (100 µm high) was used for collagen loading. The cell channel (50 µm high) was used for cell loading. The linear LPS concentration gradient was generated in fve migration channels (50 µm high). The directional velocities of healthy, general sepsis, severe sepsis, and septic shock groups were decreased with the seriousness of sepsis. Reprinted from Ref. [33], Copyright (2016), with permission from Elsevier; b the donut-shape neutrophil chemotaxis assay and characterization of neutrophil chemotaxis of adults, term neonates, and preterm neonates. Reprinted from Ref. [92] Copyright (2017), with permission from Elsevier; c the microfuidic method for phenotyping asthma patients. For asthmatic and nonasthmatic patients, no statistically signifcant diference in neutrophil migration speed (i) and chemotactic index (ii); for asthmatic patients (n 23), neutrophil chemotaxis velocity (iii) is lower than nonasthmatic patients (n 11). *P 0.02. (Figure reproduced from Ref. [39]) = = = of preterm neonates at baseline, and compared to adults, environment for neutrophils by changing the materials both term and preterm neonates had decreased neu- and the structure of the devices. Meanwhile, the stud- trophil velocity. In the presence of LPS stimulation, the ies of integrated system have been increased and had number of spontaneously migrating neutrophils of pre- the potential for clinical applications, especially some term neonates was reduced compared to term neonates researches aimed at how to achieve chemotaxis assay and adults. directly from the whole blood. However, there are still Based on the KOALA platform, Sackmann et al. [39] some challenging issues of microfuidic-based neutro- developed a microfuidic-based handheld diagnos- phil studies. Te major challenges are to precisely control tic device to discriminate asthma from allergic rhinitis the concentration gradients of chemotactic agents and to based on the patient’s neutrophil chemotactic function reduce the impact of the fuid shear forces on neutrophils. (Fig. 9c). It was determined that neutrophils from asth- Te control of other elements of the microenvironment, matic patients (n = 23) migrated signifcantly slowly such as pH and temperature, is still unstable. Te integra- toward the chemoattractant compared with nonasth- tion of the devices is low with lacking of the integration matic patients(n = 11) (P = 0.02). Te device needed low of signal acquisition and processing modules. Automated requirements of sample volume and the difusion of lami- analytical systems are needed for clinical and commercial nar fuidics provided precise control microenvironment application. Te analysis and information extraction of of cells. the results obtained from neutrophil chemotaxis assays and further modeling of the neutrophil chemotaxis pro- Conclusion cess and mechanism analysis are issues to be considered. In summary, we reviewed the recent developments in With the development of microfuidic technology, neu- microfuidic-based neutrophil chemotaxis studies with trophil chemotaxis studies will be developed in the direc- diferent gradients generation methods. Te integrated tion of multi-functions, easy-operation, and integration. microfuidic devices based on various functions were Microfuidic technology will become a useful technical also discussed. By taking the advantages of microfuid- tool to study the mechanism of neutrophil chemotaxis ics, microfuidic devices have been a popular tool for cell and microfuidic-based neutrophil chemotaxis devices chemotaxis studies. Te devices can provide stable, con- will have more application not only in basic researches trolled, and complex profles gradients and a shear-free but also in clinical applications. Zhao et al. J Transl Med (2020) 18:168 Page 17 of 19

6. Kang SK, Shin IS, Ko MS, Jo JY, Ra JC. Journey of mesenchymal stem cells for homing: strategies to enhance efcacy and safety of stem cell therapy. Abbreviations Stem Cells Int. 2012;2012:342968. bMFA: Bioinspired microfuidic assay; µ-TAS: Micro-total analysis system; 7. Saxena N, Mogha P, Dash S, Majumder A, Jadhav S, Sen S. Matrix PDMS: Polydimethlsiloxane; µC3: Microfuidic competitive chemotaxis-chip; elasticity regulates mesenchymal stem cell chemotaxis. J Cell Sci. CGG​: Concentration gradient generator; MtLN3: Metastatic rat mammary 2018;131(7):jcs211391. adenocarcinoma cells; µFCD: Microfuidic chemotaxis device; TEM: Transen- 8. Hauert AB, Martinelli S, Marone C, Niggli V. Diferentiated HL-60 cells are dothelial migration; UCB: Umbilical cord blood; WBLC: Whole blood loading a valid model system for the analysis of human neutrophil migration and chamber; FCCs: Focal chemotaxis chambers; UMCAS: USB microscope-based chemotaxis. Int J Biochem Cell Biol. 2002;34(7):838–54. microfuidic chemotaxis analysis system; PMNs: Polymorphonuclear neutro- 9. Schwager J, Bompard A, Weber P, Raederstorf D. Ascorbic acid modulates phils; TLR4: Toll-like receptor; MSCs: Mesenchymal stem cells; IL-6: Interleu- cell migration in diferentiated HL-60 cells and peripheral blood leuko- kin-6; PDGF: Platelet derived growth factor; ECM: Extracellular matrix; DMSO: cytes. Mol Nutr Food Res. 2015;59(8):1513–23. Dimethyl sulfoxide; PMN: Polymorphonuclear cells; TSP: Thrombospondin; 10. Mansfeld P, Suchard SJJI. Thrombospondin promotes both chemot- RA: Retinoic acid; l-AA: l-ascorbic acid; GPCRs: G protein-coupled receptors; axis and haptotaxis in neutrophil-like HL-60 cells. Am Assoc Immnol. PMMA: Polymethyl methacrylate; PC: Polycarbonate; MS2: Microfuidic device 1993;150(5):1959–70. and smartphone. 11. Muller WA. Leukocyte–endothelial-cell interactions in leukocyte transmigration and the infammatory response. Trends Immunol. Acknowledgements 2003;24(6):326–33. Not applicable. 12. Kolaczkowska E, Kubes P. Neutrophil recruitment and function in health and infammation. Nat Rev Immunol. 2013;13(3):159–75. Authors’ contributions 13. Petri B, Sanz MJ. Neutrophil chemotaxis. Cell Tissue Res. ZWJ and ZHP wrote the paper. LMX and HCJ reviewed and edited the manu- 2018;371(3):425–36. script. All authors read and approved the fnal manuscript. 14. Tweedy L, Susanto O, Insall RH. Self-generated chemotactic gradients- cells steering themselves. Curr Opin Cell Biol. 2016;42:46–51. Funding 15. Heinrich D, Barnett R, Tweedy L, Insall R, Stallforth P, Winckler T. The The authors acknowledge funding from the National Key Research and chemoattractant glorin is inactivated by ester cleavage during early Development Program of China (2018YFC2001103), National Science and multicellular development of Polysphondylium pallidum. ACS Chem Biol. Technology Major Project of China (2018ZX09201-011), Beijing Natural Science 2018;13(6):1506–13. Foundation (61801477) and Scientifc Research Key Program of Beijing Munici- 16. Boyden S. The chemotactic efect of mixtures of antibody and antigen on pal Commission of Education (KZ201810025041). polymorphonuclear leucocytes. J Exp Med. 1962;1(115):453–66. 17. Zicha D, Dunn G, Brown AF, Zicha D, Dunn GA, Brown AF. A new direct- Availability of data and materials viewing chemotaxis chamber. J Cell Sci. 1991;99(Part 4):769–75. Not applicable. 18. Muinonen-Martin AJ, Veltman DM, Kalna G, Insall RH. An improved cham- ber for direct visualisation of chemotaxis. PLoS ONE. 2010;5(12):e15309. Ethics approval and consent to participate 19. Wu J, Wu X, Lin F. Recent developments in microfuidics-based chemot- Not applicable. axis studies. Lab Chip. 2013;13(13):2484–99. 20. Coghill PA, Kesselhuth EK, Shimp EA, Khismatullin DB, Schmidtke DW. Consent for publication Efects of microfuidic channel geometry on leukocyte rolling assays. Written informed consent for publication was obtained from all participants. Biomed Microdevices. 2013;15(1):183–93. 21. Lin F. Chapter 15. A microfuidics-based method for chemoattractant Competing interests gradients. Methods Enzymol. 2009;461:333–47. The authors declare that they have no competing interests. 22. Li Jeon N, Baskaran H, Dertinger SK, Whitesides GM, Van de Water L, Toner M. Neutrophil chemotaxis in linear and complex gradients of Author details interleukin-8 formed in a microfabricated device. Nat Biotechnol. 1 Institute of Microelectronics, Chinese Academy of Sciences, 3 Beitucheng 2002;20(8):826–30. West Road, Beijing 100029, China. 2 School of Future Technology, University 23. Kim D, Haynes CL. On-chip evaluation of neutrophil activation and of Chinese Academy of Sciences, Beijing, China. 3 Cerebrovascular Diseases neutrophil-endothelial cell interaction during neutrophil chemotaxis. Research Institute, Xuanwu Hospital of Capital Medical University, 45 Chang- Anal Chem. 2013;85(22):10787–96. chun Street, Beijing, China. 24. Campbell K, Groisman A. Generation of complex concentration profles in microchannels in a logarithmically small number of steps. Lab Chip. Received: 5 February 2020 Accepted: 9 April 2020 2007;7(2):264–72. 25. Wu J, Hillier C, Komenda P, Lobato-de-Faria R, Levin D, Zhang M, et al. A microfuidic platform for evaluating neutrophil chemotaxis induced by sputum from COPD patients. PLoS ONE. 2015;10(5):e0126523. 26. Yang K, Wu J, Zhu L, Liu Y, Zhang M, Lin F. An all-on-chip method for References rapid neutrophil chemotaxis analysis directly from a drop of blood. JoVE. 1. Ridley AJ, Schwartz MA, Burridge K, Firtel RA, Ginsberg MH, Borisy G, 2017;124:e55615. et al. Cell migration: integrating signals from front to back. Science. 27. Tay HM, Dalan R, Li KHH, Boehm BO, Hou HW. A novel microdevice for 2003;302(5651):1704. rapid neutrophil purifcation and phenotyping in type 2 diabetes mel- 2. Nathan C. Neutrophils and immunity: challenges and opportunities. Nat litus. Small. 2018;14(6):1702832. Rev Immunol. 2006;6(3):173–82. 28. Mao H, Cremer PS, Manson MD. A sensitive, versatile microfuidic assay for 3. Mantovani A, Cassatella MA, Costantini C, Jaillon S. Neutrophils in the bacterial chemotaxis. Proc Natl Acad Sci. 2003;100(9):5449. activation and regulation of innate and adaptive immunity. Nat Rev 29. Lee JSH, Hu Y, Li D. Electrokinetic concentration gradient genera- Immunol. 2011;11(8):519–31. tion using a converging–diverging microchannel. Anal Chim Acta. 4. Petrie Aronin CE, Zhao YM, Yoon JS, Morgan NY, Prustel T, Germain RN, 2005;543(1–2):99–108. et al. Migrating myeloid cells sense temporal dynamics of chemoattract- 30. Irimia D, Charras G, Agrawal N, Mitchison T, Toner M. Polar stimula- ant concentrations. Immunity. 2017;47(5):862-74 e3. tion and constrained cell migration in microfuidic channels. Lab Chip. 5. Zigmond SH. Ability of polymorphonuclear leukocytes to orient in gradi- 2007;7(12):1783–90. ents of chemotactic factors. J Cell Biol. 1977;75(2 Pt 1):606–16. 31. Zhang W, He Z, Yi L, Mao S, Li H, Lin JM. A dual-functional microfuidic chip for on-line detection of interleukin-8 based on rolling circle amplif- cation. Biosens Bioelectron. 2018;15(102):652–60. Zhao et al. J Transl Med (2020) 18:168 Page 18 of 19

32. Saadi W, Rhee SW, Lin F, Vahidi B, Chung BG, Jeon NL. Generation of 57. Jung W, Han J, Choi J-W, Ahn CH. Point-of-care testing (POCT) diagnostic stable concentration gradients in 2D and 3D environments using a systems using microfuidic lab-on-a-chip technologies. Microelectron microfuidic ladder chamber. Biomed Microdevices. 2007;9(5):627–35. Eng. 2015;132:46–57. 33. Lu X, Lv C, Qu Y, Luo Y, Lin B, Zhan L, et al. Sepsis-induced impair- 58. Grigolato F, Egholm C, Impellizzieri D, Arosio P, Boyman O. Establishment ment of neutrophil chemotaxis on a microfuidic chip. Immunol Lett. of a scalable microfuidic assay for characterization of population-based 2016;173:55–60. neutrophil chemotaxis. Allergy. 2020. https​://doi.org/10.1111/all.14195​. 34. Kim T, Pinelis M, Maharbiz MM. Generating steep, shear-free gra- 59. Hattori K, Sugiura S, Kanamori T. Generation of arbitrary monotonic dients of small molecules for cell culture. Biomed Microdevices. concentration profles by a serial dilution microfuidic network com- 2009;11(1):65–73. posed of microchannels with a high fuidic-resistance ratio. Lab Chip. 35. VanDersarl JJ, Xu AM, Melosh NA. Rapid spatial and temporal 2009;9(12):1763–72. controlled signal delivery over large cell culture areas. Lab Chip. 60. Zhou Y, Wang Y, Mukherjee T, Lin Q. Generation of complex concentration 2011;11(18):3057–63. profles by partial difusive mixing in multi-stream laminar fow. Lab Chip. 36. Kim D, Lokuta MA, Huttenlocher A, Beebe DJ. Selective and tunable 2009;9(10):1439–48. gradient device for cell culture and chemotaxis study. Lab Chip. 61. Bourne HR, Weiner O. Cell polarity: a chemical compass. Nature. 2009;9(12):1797–800. 2002;419(6902):21. 37. Chung HH, Chan CK, Khire TS, Marsh GA, Clark A Jr, Waugh RE, et al. 62. Aranyosi AJ, Wong EA, Irimia D. A neutrophil treadmill to decouple spatial Highly permeable silicon membranes for shear free chemotaxis and and temporal signals during chemotaxis. Lab Chip. 2015;15(2):549–56. rapid cell labeling. Lab Chip. 2014;14(14):2456–68. 63. Walsh DI 3rd, Lalli ML, Kassas JM, Asthagiri AR, Murthy SK. Cell chemotaxis 38. Zhou Y, Lin Q. Microfuidic fow-free generation of chemical concentra- on paper for diagnostics. Anal Chem. 2015;87(11):5505–10. tion gradients. Sens Actuators B Chem. 2014;190:334–41. 64. Weckmann M, Becker T, Nissen G, Pech M, Kopp MV. SiMA: a simpli- 39. Sackmann EK, Berthier E, Schwantes EA, Fichtinger PS, Evans MD, fed migration assay for analyzing neutrophil migration. Cytometry A. Dziadzio LL, et al. Characterizing asthma from a drop of blood using 2017;91(7):675–85. neutrophil chemotaxis. Proc Natl Acad Sci USA. 2014;111(16):5813–8. 65. Boneschansker L, Jorgensen J, Ellett F, Briscoe DM, Irimia D. Convergent 40. Abhyankar VV, Toepke MW, Cortesio CL, Lokuta MA, Huttenlocher and divergent migratory patterns of human neutrophils inside microfu- A, Beebe DJ. A platform for assessing chemotactic migration idic mazes. Sci Rep. 2018;8(1):1887. within a spatiotemporally defned 3D microenvironment. Lab Chip. 66. Butler KL, Ambravaneswaran V, Agrawal N, Bilodeau M, Toner M, Tompkins 2008;8(9):1507–15. RG, et al. Burn injury reduces neutrophil directional migration speed in 41. Wu X, Newbold MA, Haynes CL. Recapitulation of in vivo-like neutro- microfuidic devices. PLoS ONE. 2010;5(7):e11921. phil transendothelial migration using a microfuidic platform. Analyst. 67. Keenan TM, Hsu C-H, Folch A. Microfuidic, “jets” for generating steady- 2015;140(15):5055–64. state gradients of soluble molecules on open surfaces. Appl Phys Lett. 42. Haessler U, Kalinin Y, Swartz MA, Wu M. An agarose-based microfuidic 2006;89(11):114103. platform with a gradient bufer for 3D chemotaxis studies. Biomed Micro- 68. Keenan TM, Frevert CW, Wu A, Wong V, Folch A. A new method for devices. 2009;11(4):827–35. studying gradient-induced neutrophil desensitization based on an open 43. Sackmann EK, Berthier E, Young EW, Shelef MA, Wernimont SA, Hutten- microfuidic chamber. Lab Chip. 2010;10(1):116–22. locher A, et al. Microfuidic kit-on-a-lid: a versatile platform for neutrophil 69. Yang K, Peretz-Soroka H, Wu J, Zhu L, Cui X, Zhang M, et al. Fibroblast chemotaxis assays. Blood. 2012;120(14):e45–53. growth factor 23 weakens chemotaxis of human blood neutrophils in 44. Schaf UY, Xing MM, Lin KK, Pan N, Jeon NL, Simon SI. Vascular mimetics microfuidic devices. Sci Rep. 2017;7(1):3100. based on microfuidics for imaging the leukocyte–endothelial infamma- 70. Yang K, Wu J, Xu G, Xie D, Peretz-Soroka H, Santos S, et al. A dual-docking tory response. Lab Chip. 2007;7(4):448–56. microfuidic cell migration assay (D(2)-Chip) for testing neutrophil 45. Soroush F, Zhang T, King DJ, Tang Y, Deosarkar S, Prabhakarpandian B, chemotaxis and the memory efect. Integr Biol (Camb). 2017;9(4):303–12. et al. A novel microfuidic assay reveals a key role for protein kinase C 71. Boribong BP, Lenzi MJ, Li L, Jones CN. Super-low dose lipopolysaccharide delta in regulating human neutrophil-endothelium interaction. J Leukoc dysregulates neutrophil migratory decision-making. Front Immunol. Biol. 2016;100(5):1027–35. 2019;10:359. 46. Lamberti G, Prabhakarpandian B, Garson C, Smith A, Pant K, Wang B, 72. Hamza B, Wong E, Patel S, Cho H, Martel J, Irimia D. Retrotaxis of human et al. Bioinspired microfuidic assay for in vitro modeling of leukocyte- neutrophils during mechanical confnement inside microfuidic channels. endothelium interactions. Anal Chem. 2014;86(16):8344–51. Integr Biol (Camb). 2014;6(2):175–83. 47. Simon SI, Sarantos MR, Green CE, Schaf UY. Leucocyte recruitment under 73. Kim BJ, Wu M. Microfuidics for mammalian cell chemotaxis. Ann Biomed fuid shear: mechanical and molecular regulation within the infamma- Eng. 2011;40(6):1316–27. tory synapse. Clin Exp Pharmacol Physiol. 2009;36(2):217–24. 74. Mosadegh B, Huang C, Park JW, Shin HS, Chung BG, Hwang S-K, et al. 48. Yang K, Wu J, Peretz-Soroka H, Zhu L, Li Z, Sang Y, et al. Mkit: a cell Generation of stable complex gradients across two-dimensional surfaces migration assay based on microfuidic device and smartphone. Biosens and three-dimensional gels. Langmuir. 2007;23(22):10910–2. Bioelectron. 2018;15(99):259–67. 75. Cheng S-Y, Heilman S, Wasserman M, Archer S, Shuler ML, Wu M. A 49. Manz A, Graber N, Widmer HM. Miniaturized total chemical analysis hydrogel-based microfuidic device for the studies of directed cell migra- systems: a novel concept for chemical sensing. Sens Actuators B Chem. tion. Lab Chip. 2007;7(6):763–9. 1990;1(1):244–8. 76. Ahmed T, Shimizu TS, Stocker R. Bacterial chemotaxis in linear and nonlin- 50. Thorsen T, Maerkl SJ, Quake SR. Microfuidic large-scale integration. Sci- ear steady microfuidic gradients. Nano Lett. 2010;10(9):3379–85. ence (New York, NY). 2002;298(5593):580–4. 77. Keenan TM, Folch A. Biomolecular gradients in cell culture systems. Lab 51. Chen X, Cui D-F. Microfuidic devices for sample pretreatment and appli- Chip. 2008;8(1):34–57. cations. Microsyst Technol. 2009;15(5):667–76. 78. Sip CG, Bhattacharjee N, Folch A. Microfuidic transwell inserts for 52. Liu L, Chen K, Xiang N, Ni Z. Dielectrophoretic manipulation of nanoma- generation of tissue culture-friendly gradients in well plates. Lab Chip. terials: a review. Electrophoresis. 2019;40(6):873–89. 2014;14(2):302–14. 53. Lee CY, Chang CL, Wang YN, Fu LM. Microfuidic mixing: a review. Int J Mol 79. Joanne Wang C, Li X, Lin B, Shim S, Ming GL, Levchenko A. A microfu- Sci. 2011;12(5):3263–87. idics-based turning assay reveals complex growth cone responses to 54. Jia Y, Ni Z, Sun H, Wang C. Microfuidic approaches toward the isolation integrated gradients of substrate-bound ECM molecules and difusible and detection of exosome nanovesicles. IEEE Access. 2019;7:45080–98. guidance cues. Lab Chip. 2008;8(2):227–37. 55. Cho S, Kang DK, Choo J, de Mello AJ, Chang SI. Recent advances in micro- 80. Kim D, Haynes CL. Neutrophil chemotaxis within a competing gradient of fuidic technologies for biochemistry and molecular biologys. BMB Rep. chemoattractants. Anal Chem. 2012;84(14):6070–8. 2011;44(11):705–12. 81. Wu X, Kim D, Young AT, Haynes CL. Exploring infammatory disease drug 56. Srinivasan V, Pamula VK, Fair RB. An integrated digital microfuidic lab-on- efects on neutrophil function. Analyst. 2014;139(16):4056–63. a-chip for clinical diagnostics on human physiological fuids. Lab Chip. 82. Mathias B, Mira JC, Rehfuss JP, Rincon JC, Ungaro R, Nacionales DC, 2004;4(4):310–5. et al. LPS stimulation of cord blood reveals a newborn-specifc Zhao et al. J Transl Med (2020) 18:168 Page 19 of 19

neutrophil transcriptomic response and cytokine production. Shock. 91. Jones CN, Hoang AN, Martel JM, Dimisko L, Mikkola A, Inoue Y, et al. 2017;47(5):606–14. Microfuidic assay for precise measurements of mouse, rat, and 83. Kotz KT, Xiao W, Miller-Graziano C, Qian W-J, Russom A, Warner EA, et al. human neutrophil chemotaxis in whole-blood droplets. J Leukoc Biol. Clinical microfuidics for neutrophil genomics and proteomics. Nat Med. 2016;100(1):241–7. 2010;16(9):1042–7. 92. Raymond SL, Mathias BJ, Murphy TJ, Rincon JC, Lopez MC, Ungaro R, 84. Agrawal N, Toner M, Irimia D. Neutrophil migration assay from a drop of et al. Neutrophil chemotaxis and transcriptomics in term and preterm blood. Lab Chip. 2008;8(12):2054–61. neonates. Transl Res. 2017;190:4–15. 85. Moussavi-Harami SF, Pezzi HM, Huttenlocher A, Beebe DJ. Simple micro- 93. Raymond SL, Hawkins RB, Murphy TJ, Rincon JC, Stortz JA, Lopez MC, et al. fuidic device for studying chemotaxis in response to dual gradients. Impact of toll-like receptor 4 stimulation on human neonatal neutrophil Biomed Microdevices. 2015;17(3):9955. spontaneous migration, transcriptomics, and cytokine production. J Mol 86. Moussavi-Harami SF, Mladinich KM, Sackmann EK, Shelef MA, Starnes TW, Med (Berl). 2018;96(7):673–84. Guckenberger DJ, et al. Microfuidic device for simultaneous analysis of 94. Wang X, Irimia D. Neutrophil chemotaxis in one droplet of blood using neutrophil extracellular traps and production of reactive oxygen species. microfuidic assays. Methods Mol Biol. 2018;1749:351–60. Integr Biol (Camb). 2016;8(2):243–52. 95. Wu J, Ouyang L, Wadhawan N, Li J, Zhang M, Liao S, et al. A compact 87. Jones CN, Dalli J, Dimisko L, Wong E, Serhan CN, Irimia D. Micro- microfuidic system for cell migration studies. Biomed Microdevices. fuidic chambers for monitoring leukocyte trafcking and human- 2014;16(4):521–8. ized nano-proresolving medicines interactions. Proc Natl Acad Sci. 96. Yang K, Peretz-Soroka H, Liu Y, Lin F. Novel developments in mobile sens- 2012;109(50):20560. ing based on the integration of microfuidic devices and smartphones. 88. Hoang AN, Jones CN, Dimisko L, Hamza B, Martel J, Kojic N, et al. Measur- Lab Chip. 2016;16(6):943–58. ing neutrophil speed and directionality during chemotaxis, directly from a droplet of whole blood. Technology. 2013;01(01):49–57. 89. Jones CN, Hoang AN, Dimisko L, Hamza B, Martel J, Irimia D. Microfuidic Publisher’s Note platform for measuring neutrophil chemotaxis from unprocessed whole Springer Nature remains neutral with regard to jurisdictional claims in pub- blood. J Vis Exp. 2014;3(88):e51215. lished maps and institutional afliations. 90. Hamza B, Irimia D. Whole blood human neutrophil trafcking in a microfuidic model of infection and infammation. Lab Chip. 2015;15(12):2625–33.

Ready to submit your research ? Choose BMC and benefit from:

• fast, convenient online submission • thorough peer review by experienced researchers in your field • rapid publication on acceptance • support for research data, including large and complex data types • gold Open Access which fosters wider collaboration and increased citations • maximum visibility for your research: over 100M website views per year

At BMC, research is always in progress.

Learn more biomedcentral.com/submissions