**1. Introduction**

In the field of endocytosis and exocytosis, a precise manipulation of a cell was required, especially for the nanomedicine injection, intracellular signaling pathway, and pathogenic research, which holds a great potential in the detection and treatment of complex diseases [1]. Thus, increasing attentions have been paid to the dynamic manipulation of cells in a fluid, which has been proved to be crucial in cell growth [2, 3], differentiation [4–6], drug delivery [7–9], mechanical force transduction [10, 11], etc. Up to now, diverse techniques have shown invaluable performance in the multifunctional manipulation of cells, including dielectrophoretic [12, 13], magnetic field [14, 15], and mechanical force [4, 16, 17]. Nevertheless, these strategies face the challenges of potential sample heating and high power consumption and required attachment of electric/magnetic nanoparticles [18, 19]. Besides, the cells were required to be confined on an elaborated substrate, which was usually not reusable in these techniques [20, 21].

Since first introduced by A. Ashkin, optical tweezers have motivated many intriguing advances in interdisciplinary applications, such as micro/nanophotonics, biphotonics, and biomedicine [22–24]. The most commonly used optical manipulation tool is the conventional optical tweezers (COTs), which uses high numerical aperture objectives to focus the free-space laser beam [25, 26]. After the transfer of photon momentum, the cell will suffer from the optical force, which can be

divided into optical scattering force and gradient force, with the direction pointed to the optical propagating direction and beam focus, respectively. The magnitude of optical force was ranged from femtonewton to nanonewton, and thus, it was the ideal chose to measure the response of biological and macromolecular system [27]. Compared to other microscopic techniques, for example, electric filed, magnetic field, and acoustic method, optical tweezers have the great potential of high precision, high flexibility, noncontact, and wide manipulation range. However, COTS faces certain challenges, such as bulky structure, limited integration, and diffraction for nanoparticles [28]. Thus, researchers have developed various schemes of optical tweezers to extend the potential application scenario, including holographic optical tweezers (HOTs), surface plasmon-based optical tweezers (SPOTs), and fiber probe-based optical tweezers (FP).

For HOTs, the optical tweezer system was inserted with a diffractive beam splitter. Then the beam wavefront can be further sculptured, which can manipulate multiple cells simultaneously [29–31]. Nevertheless, appropriate algorithms are required to be elaborately designed for achieving a specific pattern of cells. In addition, the complicated optical system was consisted of dichroic mirrors, spatial light modulators, and high-numerical-aperture focusing objectives. The working distance of objectives limits the depth at which cells can be manipulated in the cell suspensions. Besides HOTs, surface plasmon-based optical tweezers (SPOTs) have also been developed to manipulate cells with high trapping stability and retaining ability, especially for the sub-microsized cells and biological molecules [32, 33]. However, once the substrates designed, the cell pattern is fixed without a flexibility to adjust and transport the cell chain dynamically. Moreover, the organization of cell pattern is also limited at a specific depth of cell suspension. Thus, there is a great need of developing a new strategy to perform the multifunctional manipulation of cells.
