How far should scientists be allowed to go in creating things that resemble primitive human brains, hearts, and even human embryos?

That's the question being asked by a group of Harvard scientists who are doing exactly that in their labs. They're using stem cells, genetics and other new biological engineering techniques to create tissues, primitive organs and other living structures that mimic parts of the human body.

Their concern is that they and others doing this type of "synthetic biology" research might be treading into disturbing territory.

"We don't know where this going to go," says John Aach, a lecturer in genetics at Harvard Medical School. "This is just the beginning of this field."

Aach helped write a paper in the journal eLife, published Tuesday, calling for an international effort to establish guidelines for this provocative area of research.

While all this may sound like something out of Frankenstein, the goal is to find new ways to decipher the mysteries of human biology and to discover novel treatments for health problems ranging from infertility to aging.

"We want to understand biology of natural human development and disease and come up with ways of addressing the problems of disease," Aach says. "The more precisely you can make something that is like a tissue or a system of tissues in a dish, the easier it is to experiment on it."

But in the process of conducting their experiments, Aach and his lab colleagues realized scientists might cross disturbing ethical lines.

For example, scientists could create primitive beating hearts and primordial brains.

"How much moral concern should we have for these things? If it has a brain that doesn't look like a human brain, but it operates like one, it could still feel pain," Aach says.

Some scientists have already started creating entities that resemble the very early stages of human embryos. Scientists use different names to describe them. They're sometimes called "embryoids," but Aach's group has dubbed them "SHEEFs" — synthetic human entities with embryo-like features.

In some of these experiments, researchers have seen early signs of the formation of the "primitive streak," which is the beginning of a central nervous system and, potentially, the ability to sense pain.

That work raises the prospect that the experiments might violate the 14-day rule, which has been in place for decades to avoid raising too many ethical concerns about experimenting on human embryos. Two weeks into embryonic development is usually when the primitive streak begins to appear.

But Aach and his colleagues argue that the 14-day rule, which is a guideline in the United States and law in some other countries, has become outdated by this latest generation of experiments.

It's based on the predictable, linear development of a normal human embryo. But the new synthetic biology techniques do not necessarily follow that road map.

"The primitive streak was like a stop sign," Aach says. "If you stopped there you would never get a brain. You would never get a heart. You would never get something that would be morally concerning."

"But now with these tissue engineering and stem cell techniques you can simply go around that," Aach says. "You could create something at a point beyond that. It might become sentient."

It's also possible that some day these embryoids could become so much like a normal human embryo that they could actually be used to create a baby.

So, in essence, "you've gone off-road," Aach says. "With these synthetic tissues there's no longer one highway of development. A stop sign is no longer good enough."

The ethical concerns are not just limited to structures that resemble embryos, Aach says.

As a result, he and the co-authors of the report say new guidelines are needed to replace that clear stop sign with something that's more like a guardrail or fence that will keep scientists from inadvertently steering into ethically troubling terrain.

"What we're proposing is, instead of doing stop signs, we get these perimeter fences — where there's an agreement that there's an area of concern," Aach says.

For example, scientists, philosophers, bioethicists and others may reach a consensus that "we can't make a brain that will allow it to feel pain" or "we can't make something like a heart — but we can make up to it," Aach says, "as long as it doesn't start beating."

Others scientists praised the researchers for raising these tough issues early.

"I absolutely support this," Magdelena Zernicka-Goetz tells Shots in an email; she is doing similar research at the University of Cambridge in the United Kingdom. "The time is right to begin discussion of these issues in a forum that includes scientists and has a wide representation of society," Zernicka-Goetz says.

Some bioethicists also welcomed a debate about these issues.

"I really have to give them credit for raises these issues proactively," says Insoo Hyun, a Case Western Reserve University bioethicist. "Our current standards for oversight and ethics are not adequate to capture this new area of science."

But it could be difficult to draw the line in some cases, Hyun notes, such as in experiments aimed at developing treatments for pain or those aiming to understanding the heart better.

"Those types of experiments may be exactly the point of why you'd want to create a synthetic entity that does have some kind of pain sensation, or that has some sort of neural network, or has some sort of heart beat, if that's actually the body system you want to study," Hyun says.

And, he says, there may be some experiments people find disturbing on a visceral level.

"Some people may just find that the experiments are just kind of creepy," Hyun says. "There may be some people concerned about scientists taking the research too far, creating entities in the dish that are quasi-human — and [that they] de-value life in the process."

Ali Brivanlou, an embryologist at the Rockefeller University who is conducting some of the most advanced work in this area, also says he welcomes a debate. But worried about putting too many limitations on the research.

Copyright 2017 NPR. To see more, visit http://www.npr.org/.