Hopf bifurcation analysis of nonlinear HIV infection model and the effect of delayed immune response with drug therapies

A mathematical model of HIV infection with the combination of drug therapy including cytotoxic T-lymphocyte (CTL) and the antibody immune response is examined. The threshold value represented as the basic reproduction ratio R0 is derived. This reveals that R0 < 1 is locally asymptotically stable in the viral free steady state, and the infected steady state condition remains locally asymptotically stable with R0 > 1 in the absence of a delay in the immune response. Moreover, the existence of Hopf bifurcation with CTL response delay is demonstrated. The estimation of delay length is used to maintain stability. Numerical simulations are implemented to explain the mathematical results.


Introduction
Human immunodeficiency virus (HIV) is a simple form of virus that affects the human body, and it primarily targets CD4 + T-cells. Viruses, in general, are the infectious creatures that have no reproductive ability by themselves. They depend on a host for replication to begin. HIV viruses carry a copy of their RNA that needs to first be duplicated into DNA [1]. Furthermore, once the DNA virus has been copied into these host cells, new virus particles must also be assembled at the host cell surface [2]. The maturing for these new particles might occur gradually without impacting that host cell, or it quickly erupts and destroys the host cell [3,4]. Antiretroviral therapy (ART) would intend to inhibit that activity for different HIV viral proteins and consequently hinder the viral replication cycle [5,6].The drug based on the inhibitor of reverse transcriptase (RTI) is an obstacle to the reverse transcription of viral DNA and prevents a productive infection of the host cell. Protease inhibitor (PI) drugs aim to suppress protease activity and prevent the development of mature viruses. The combination of reverse transcriptase inhibitor and protease inhibitor is more effective to treat HIV patients.
Mathematical modeling can contribute to the study of antiviral infection treatment and increase knowledge about the virus transmission rate [7,8]. Numerous models of viral infections are discussed in the literature for HIV dynamics [9]. The immune system CTL plays a vital role in protecting against HIV and this ability prevents the reproduction of HIV [10]. Hence many mathematical models examine viral infection in conjunction with CTL responses [11][12][13]. To resist viral infection, both CTL immune response and strong antibody neutralizing are necessary for an effective vaccine [14,15].
In [16] Wodarz suggested a model depicting the relationship between CTL and antibody immune responses. The global analysis for this model was provided by Yousfi et al. [17].
Yan and Wang included the intracellular delay in the cell infected model and analyzed the global dynamics effect of the delay [18].
The mathematical study is needed for an integrated view of the dynamics of viruses for delayed models [19][20][21]. Several models have been suggested with a time lag in triggering the immune reaction when the body is infected with the virus, which is called immunological delay [22][23][24].
In [25] Dubey et al. described the dynamics of infection by HIV involving available drug therapies, reverse transcriptase inhibitor (RTI), and protease inhibitor (PI) that includes uninfected CD4 + T-cells x(t), infected cells y(t), free virus v(t), and CTL immune response C(t), and antibody A(t). The non-linear differential equations of the first order are as follows: where δ 0 = δ 1 0 + ω 1 α 0 μ 10 , α 1 = α 1 1 + μ 1 α 0 μ 10 . Here, λ is the production level for uninfected CD4 + T-cells, λ 0 is an uninfected CD4 + T-cells mortality rate, r is the maximum rate of proliferation, β is the infection rate of uninfected cells by the virus, x m = x max is their carrying capacity, η r is the rate of inhibitor reverse transcriptase therapy to destroy infected cells, and η p is the rate at which the protease inhibitor therapy blocks the infection, δ 1 0 is the infected cell mortality rate, N is the number of particles of virus produced by the cell infected, δ 1 is the rate of virus removed from the body by natural factors, α 0 is the rate of produced CTL immune response, μ 10 is the deplete rate of CTL. α 1 1 is the stimulated rate of CTL, μ 1 is the rate at which CTL interact with the infected cells, ω 1 is the level of CTL interaction removed from infected cells. μ 2 is the rate at which antibody gets stimulated, μ 20 is the deplete rate of antibody due to increase in virus particles. ω 2 is the reduced rate of virus with antibody interaction.
In this study, we implement a time delay to illustrate the CTL response model. The generation of CTL due to antigenic stimulation requires a period of time τ and the response to CTL in time t that depends on an antigen population in the previous time tτ for a time lag τ > 0. The model suggested by incorporating immune response delay in CTL is as follows: The primary objective is to analyze the impact of immune reaction delay in the dynamics of (2) together with reverse transcriptase inhibitor and protease inhibitor and show the impact on the stability of the delay in the immune response.
We start with some notes that are used in the sequel to our study.
The solution (x(t), y(t), v(t), C(t), A(t)) is easy to recognize from the basic theory of differential functions of (2) with the initial conditions as mentioned above which exist for all t ≥ 0 and are unique [26][27][28]. It can be shown that this solution exists for all t > 0 and stays nonnegative. In fact, if x(0) > 0, then x(t) > 0 for all t > 0. The same argument is true for y, v, C, and A components. Hence, the interior R 5 + is invariant for system (2). The basic reproduction ratio R 0 of model (2) is defined by

Stability of viral free steady state (I 0 )
Theorem 2.1 If R 0 < 1 then the viral free steady state I 0 is locally asymptotically stable.
Proof The viral free steady state of system (2) is I 0 (x 0 , 0, 0, 0, 0), where The characteristic equation of I 0 is given by Thus, the eigenvalues of (3) have negative real parts. Hence the viral free steady state of (2) is locally asymptotically stable.

Stability analysis of infected steady state (I * ) and existence of Hopf bifurcation
Theorem 3.1 If R 0 > 1 then (i) the infected steady state I * of system (2) is locally asymptotically stable for τ = 0; (ii) Time delay τ crosses the critical value τ 0 , then system (2) undergoes Hopf bifurcation at the infected steady state I * .
Proof The infected steady state I * (x * , y * , v * , C * , A * ) of (2) is given by The uninfected CD4 + T cells (x * ) can achieve a steady state value by resolving the equation The immune response (C * ) can obtain a steady state value by solving the following equation: The characteristic equation of I * is given by This is equivalent to the following equation: Case (i). If R 0 > 1 and τ = 0. If τ = 0 then equation (4) becomes Here, u 1 = k 1 + l 1 , u 2 = k 2 + l 2 , u 3 = k 3 + l 3 , u 4 = k 4 + l 4 , u 5 = k 5 + l 5 .
According to the Routh-Hurwitz condition (i) u i > 0, i = 1, 2, 3, 4, 5; ; hence all the roots of (5) have negative real parts, and therefore the infected steady state I * is locally asymptotically stable.
Case (ii). If τ = 0. If τ = 0 then the bifurcation parameter for the delay in immune response when analyzing the existence of Hopf bifurcations starting from infected steady state I * .

Theorem 4.1
If there exists a delay in 0 < τ < τ + such that B 1 τ 2 + B 2 τ < B 3 , then τ + is locally asymptotically stable, the delay length for the infected steady state I * .

Results and discussion
We exhibit some numerical assumptions with respect to our hypothetical investigation which are done in MATLAB. Using the following parameter values  (8) and (9) we obtain τ 0 = 1.102 and R 0 = 21.7348 > 1. In case that the delay moderately increases from zero, at that point the infected steady state I * is locally asymptotically stable for τ ∈ [0, τ 0 ) (by Theorem 3.1) which is outlined in Fig. 1. In any case, the delay crosses the critical value τ 0 = 1.102,the infected steady state I * loses stability and undergoes Hopf bifurcation, which is illustrated in Fig. 2. The proposed demonstration suggests that as the response of antibody often evolves as possible in free virus Figure 1 The phase diagrams of system (2) being asymptotically stable when τ = 0.9

Figure 2
The phase diagrams of system (2) undergoing Hopf bifurcation when τ = 1.2  The phase diagram demonstrates that the neutralization of the antibody regulates the infected cells as antibody neutralization increases Figure 6 The phase diagram shows that when neutralization of antibody response increases and the free virus level converges to zero, the free virus is neutralized

Conclusion
In this study, the dynamics of an HIV infection model were analyzed and described by comprising the CTL response delay according to mathematical analysis. We have shown that the reproduction ratio R 0 plays a crucial role for inferring this model dynamics. The calculation of R 0 in an infectious disease transmitted by blood or sex is more complicated. Several mathematical models have been proposed to explain the rate of spread of the infection. An account should be taken of the window period and HIV pathophysiology. Better integration of modeling into the decision-making process is desirable to improve the effectiveness of interventions and the knowledge of health professionals [32].
Both the viral free steady state and the infected steady state of the HIV model are ascertained. If R 0 < 1 then the viral free steady state is locally asymptotically stable, and if R 0 > 1 then infected steady state is locally asymptotically stable in absence of delay. Based on the model of differential delay, we evaluate Hopf bifurcation requirements using time delay as the bifurcation parameter. This shows that a positive steady state is locally asymptotically stable when the time delay is relatively small, whereas a Hopf bifurcation may cause a loss of stability as the delay increases. Using the Nyquist test, we also get that the maximum delay value to the infected steady State I * will remain asymptomatically stable. Finally, it can be concluded that an antibody immune reaction and the combination of drug efficacy are important to reduce the potential for HIV infections through numerical simulations. Maintaining stability may be the goal to prolonging patient survival, but exposes them to transmission risk. It is interesting to know the duration of the asymptotic period for subsequent consequent clinical evaluations. Dynamic HIV transmission models can provide evidence-based guidance on optimal combination implementation strategies to treat and prevent HIV/AIDS [33].