AND gates are part of logic circuits which are often used in synthetic biology (Wang et al., 2015); the idea is to integrate inputs into outputs. For an AND gate two inputs will be integrated to generate one output.
In order to establish the necessary or optimal amount of the inputs to cause the output in silico modeling is implemented. For this the following has to be true: if two inputs are given (+/+), the output is ‘ON’. In case only one input (+/- or -/+) or none (-/-) is given, the output is ‘OFF’ (Fig. 1). This system can be put into logic functions (Lomnitz & Savageau, 2016).
In our project we used this approach for a controlled expression of our output, the chimeric antigen receptor (CAR). The two inputs used to generate this output are high VEGF concentrations and hypoxic conditions or alternatively low pH and hypoxic conditions.
In case that only one of the two inputs is given, the CAR will not be expressed and the CAR T cell will not be able to bind to its target antigen.
Through the dependent expression of the CAR, we want to achieve a higher specificity to solid tumor cells, considering that both VEGF and hypoxic conditions or a low pH and hypoxia should only be given in the direct tumor microenvironment. This will lead to an improvement of the safety of immunotherapy.
Due to lowered oxygen concentrations in the tumor microenvironment the cells have to respond to hypoxia (1 % oxygen) and gene expression has to be activated. For this the cells possess HIFs (hypoxia-inducible factors), which are heterodimeric transcription factors and are composed of a hypoxia dependent alpha subunit and a constitutively expressed beta subunit (Pescador et al., 2005).
In our project we focused on HIF1, however there are also HIF2 and HIF3. Although they are regulated in a similar way, the function and distribution in the tissue varies (Pescador et al., 2005).
Under normoxic conditions HIF1A is hydroxylated and is marked by the E3 ubiquitin ligase which leads to the degradation by the proteasome. However, in hypoxic conditions HIF1A is stabilized and can heterodimerize with HIF1B. HIF1 can then bind to hypoxia response elements (HRE) in the nucleus and lead to the expression of the gene of interest (Fig. 2) (Ziello et al., 2007), which in case of the CARTELTM AND gate is the CAR.
Since HIF1A is already endogenously expressed, it is necessary to create a knockout or knockdown of the endogenous HIF1A, in order to be able to reintroduce it within the CARTELTM AND gate. Otherwise the endogenous HIF1A would be able to interfere with the CARTELTM AND gate and leading to its activation without the presence of the chosen second input.
VEGF (Vascular Endothelial Growth Factor)
VEGF is one of the chosen inputs of the CARTELTM AND gate. Normally it initiates blood coagulation but in our system it is utilized to drive the expression of HIF1A. For this the endogenous pathway of VEGF is used, which functions as follows:
VEGF-A is primarily recognized by its receptor VEGFR-2 (vascular endothelial growth factor receptor 2) which is often expressed on tumor cells (Miettinen et al., 2013).
VEGFR-2 is a receptor tyrosine kinase, leading to phosphorylation steps upon binding of its activator VEGF. Furthermore, intracellular calcium concentrations increase and the phosphatase calcineurin, which is dependent on calcium, is activated. Calcineurin is then able to dephosphorylate the transcription factor NFAT (nuclear factor of activated T cells), which is then able to enter the nucleus and activate the CTLA4 (cytotoxic T-lymphocyte-associated protein 4) promoter (Minami et al., 2013), leading to the expression of the gene of interest (Fig. 3) - in our case HIF1A.
A low pH is another possible input for the AND gate and an alternative to the VEGF input. Extracellular protons can be sensed by the receptor TDAG8 (T cell death-associated gene 8), also called GPR65 (G protein-coupled receptor 65) and lead to the activation of a G protein. The G protein activates the adenylyl cyclase, which is able to convert ATP into cyclic AMP (cAMP). cAMP can bind and activate PKA (protein kinase A), resulting in the release of a subunit that is translocated into the nucleus where it can phosphorylate the transcription factor CREB1 (cAMP-responsive element-binding protein 1). CREB1 can bind to a synthetic CRE promoter (Ausländer et al., 2014), leading to the transcription of HIF1A.